Skip to main content

Full text of "Motivation And Morale In Industry"

See other formats


= OQ 

OUP 408 16-6-64 5,000. 


Call No. G *$< .-&> \ v BSr\ Accession No. 

Author VtVVeS;Hovv;s S. 

Title \Aot\vo.V\o^ OLYvA rrxo-ccAo. (*-. 

This book should be returned on or before the date last marked below. ^*V . 



Books by Morris S. Viteles 













Preface ix 

Acknowledgments x * 


1. Industry's Core Problem in the Utilization of Manpower 3 

2. Wage-Incentive Systems 18 

3. The Worker and Wage Incentives 45 



4. Motives, Attitudes, and Incentives 65 

5. The Impact of Social Psychology upon the Investigation of 

Motivation and Morale in Industry 83 

6. Field Theory and Action Research in Group Dynamics 110 


7. The Influence of Job Preference, Size of Work Group, and 

Knowledge of Results 127 

8. Supervision, Productivity, and Morale 148 

9. Employee Participation in Decision-Making 163 

10. The Work Group as an "Informal Social Organization" 180 

11. Teamwork 195 


12. Employee-Attitude Surveys: Scope, Methods, and Guiding 

Principles 221 


13. The Employee "Looks Over" the Plant Situation 254 

14. The Assessment of Employee Morale 271 

15. The Pay Envelope 291 

16. The Need for Security 302 

17. Status, Recognition, and the Role of Supervision 319 

18. Attitudes toward the Union 333 

19. Employee and Labor Relations 359 


20. What Is on the Worker's Mind: A Summary 381 

21. Applying the Findings of Employee- Attitude Surveys 394 

22. Moulding and Modifying Employee Attitudes 414 

23. Motivation and Morale: A Challenge to Management 440 




THERE ARE three great continuing needs in industry, 
namely, (1) to increase production, (2) to promote employee satisfac- 
tion and adjustment at work, (3) to curtail industrial strife. 

The achievement of these objectives is necessary for the sound and 
profitable operation of the business enterprise. Their achievement is 
also necessary from a broad social point of view, since neither the ma- 
terial goals of our civilization nor the equally important goals of pre- 
serving and extending the less tangible benefits of our system can be 
fully attained unless industrial productivity is high, workers are satis- 
fied, and industrial harmony prevails. 

These objectives cannot be fully achieved in any single plant, or in 
industry and business as a whole, unless accurate information is avail- 
able as to the effectiveness of the various appeals or incentives which 
can be used to arouse the co-operation of individual workers. This in- 
volves an exact determination of basic motives in work, in the sense of 
evaluating the nature and strength of various wants and needs which 
require gratification through the work situation. Only with such knowl- 
edge can management arrive at a balanced and effective program of 
personnel policies and practices which will produce maximum results in 
the attainment of goals which simultaneously have significance for in- 
dustry, the worker, and society at large. 

This volume is devoted to a comprehensive description and critical 
evaluation of American and British experimental studies and attitude 
surveys bearing upon the sources of motivation and the determinants 
of morale in industry. Throughout the book emphasis is placed upon 
the application of research findings in increasing employee productivity, 
job satisfaction, and morale in an atmosphere of harmonious relations 
between management and workers both as individuals and as members 
of informally and formally organized groups. In this respect, as is par- 
ticularly apparent in the final section, the book is primarily directed to 
management and to others concerned with the organization and op- 



eration of the industrial enterprise and, more specifically, with the for- 
mulation and administration of personnel policies. 

Although directed to management, the book will, it is felt, fill the 
need of students and teachers in psychology and related fields for a 
comprehensive and systematic treatment of motivation and morale 
projected against a background of motivational theory. There are as- 
sembled in this book not only studies drawn from the widespread lit- 
erature of psychology, but also materials which are frequently not 
available to university personnel. These are considered in a broad con- 
text of related disciplines most particularly, industrial and labor re- 
lations. Because of this approach and of the nature of the problems 
covered, it is hoped that the book will be of use in courses in industrial 
management, personnel administration, industrial engineering, and the 
like, as well as in courses in the field of psychology. 

Many sources have been consulted in the preparation of this book. 
Acknowledgments to those responsible for materials cited in the text 
are to be found on pages xi-xvi. The author is also indebted to several 
of his colleagues at the University of Pennsylvania for reading portions 
of the manuscript. Dr. Malcolm G. Preston provided a number of con- 
structive ideas on the content and organization of portions of the book. 
Dr. Francis W. Irwin and Dr. J^mes C. Diggory contributed extremely 
helpful suggestions with respect to the treatment, in Part 2, of the theo- 
retical backgrounds of motivation and morale. Chapter 6, devoted to a 
discussion of field theory and action research in group dynamics, has 
benefited from review by Dr. Albert Pepitone. Considerable help was 
given by Dr. William W. Wilkinson both in the collation of source 
materials for parts of this book and in reading proof. The author ap- 
preciates the assistance of many others, including graduate assistants 
and his secretarial staff, in sharing the burden of proofreading. To his 
wife, to whom this book is dedicated, the author owes much for her 
understanding support in meeting the stresses and strains associated 
with the process of bringing forth a book. 

Philadelphia, Pa. 
May 28, 1953 


THE MATERIALS for this book have been drawn from 
many sources. The author is indebted to the editors of many journals, 
to publishers, and to other organizations, as well as to professional col- 
leagues, for permission to use these materials. 

A considerable number of figures and tables have been taken from 
The Public Opinion Index for Industry, representing a series of con- 
fidential reports on attitudes and opinion surveys prepared by the 
Opinion Research Corporation for distribution to clients on a subscrip- 
tion basis. The author desires to express appreciation to this organization 
for special permission to reproduce materials which are ordinarily avail- 
able for only limited circulation. 

The National Industrial Conference Board has given permission to 
cite data and to use quotations from special studies prepared for dis- 
tribution to its members, and from its journals, including Management 
Record and Business Record. A similar courtesy has been extended to 
the author by the American Management Association with reference 
both to reports and to journals, including Personnel and The Manage- 
ment Review, published by AMA. 

Considerable use has been made of publications issued by the Survey 
Research Center, Institute for Social Research, University of Michigan. 
In addition, the author has drawn on mimeographed reports made avail- 
able to him by the staff of the Survey Research Center. 

The Controller of Her Britannic Majesty's Stationery Office has 
granted permission to reproduce a number of figures and quotations 
from reports on investigations conducted by the Industrial Health Re- 
search Board. Frequent reference has been made to publications of the 
Industrial Relations Research Association; the Social Research Service, 
Michigan State College; the Society for the Psychological Study of So- 
cial Issues; the National Institute of Industrial Psychology of Great 
Britain, and the Industrial Welfare Division, Department of Labour 
and National Service, Commonwealth of Australia. Materials have also 



been drawn from reports prepared by the Institute for Research in Hu- 
man Relations; the Institute for Associated Research, Dartmouth Col 
lege; Central Surveys, Inc.; the Economic and Business Foundation, 
Inc.; the Bureau of National Affairs, Inc.; Fisher, Rudge 6 Neblett, Inc.; 
the Kansas City Power & Light Co.; the Detroit Edison Co.; and The 
Lincoln Electric Co. 

The author is indebted to the editors and publishers of the follow- 
ing periodicals for permission to quote materials as identified by biblio- 
graphical references scattered throughout this book: 

Advanced Management; American Journal of Sociology; American Psy- 
chologist; American Sociological Review; Annals of the American Acad- 
emy of Political and Social Science; Archives of Psychology; Canadian 
Journal of Economids and Political Science; Dun's Review; Factory Man- 
agement and Maintenance; Forbes; Fortune; Harvard Business Review; 
Human Organization; Human Relations; International Journal of Opin- 
ion and Attitude Research; Journal of Abnormal and Social Psychology; 
Journal of Applied Psychology; Journal of Consulting Psychology; Jour- 
nal of Social Issues; Journal of Social Psychology; Mill and Factory; 
Modern Industry; Modern Management; Nation's Business; Occupa- 
tions; Occupational Psychology; Personnel Journal; Personnel Psychol- 
ogy; Psychological Bulletin; Psychological Review; Public Opinion Quar- 
terly; Quarterly Journal of Experimental Psychology; Reader's Digest; 
School and Society, and Sociometry. 

Grateful acknowledgment is made to publishers and other copyright 
holders of the books listed below, from which quotations, tables, or 
figures have been drawn with permission, by special arrangement with 
the respective copyright holders. 

J. Barbash, Labor Unions in Action. Harper and Bros., 1948. 
W. V. Bingham, Psychology Today. The University of Chicago Press, 

C. A. Bird, Social Psychology. Copyright 1940, Appleton-Century- 

Crofts, Inc. 
H. Cantril (Ed.), Gauging Public Opinion. Princeton University Press, 

H. Cantril, The Psychology of Social Movements. Copyright 1941, John 

Wiley and Sons, Inc. 
H. Cantril, The "Why" of Man's Experience. Copyright 1950, The 

Macmillan Co. 
S. Chase, et al., The Social Responsibility of Management. New York 

University School of Commerce, Accounts and Finance, 1951. 
Committee on Undersea Warfare, Human Factors in Undersea War- 
fare. Washington: National Research Council, 1949. 


Current Trends in Industrial Psychology. University of Pittsburgh Press, 

E. Dale, Greater Productivity through Labor-Management Cooperation 

(Research Report No. 14). New York: American Management 

Association, 1949. 
J. Bollard, L. W. Doob, N. E. Miller, O. H. Mowrer, and R. E. Sears, 

Frustration and Aggression. Yale University Press, 1939. 
P. F. Drucker, The New Society. Harper and Bros., 1950. 
R. Dubin, Human Relations in Administration. Prentice-Hall, Inc., 


H. B. Elkind, Preventive Management. B. C. Forbes and Co., 1931. 
C. E. Evans and La V. N. Laseau, My Job Contest (Personnel Psy- 
chology Monograph No. 1). Washington: Personnel Psychology 

Inc., 1950. 

Experience with Employee Attitude Surveys, (Studies in Personnel Pol- 
icies, No. 115). New York: National Industrial Conference Board, 

Factors Affecting Employee Morale, (Studies in Personnel Policy No. 

85). New York: National Industrial Conference Board, 1947. 
E. A. Fleishman, "Leadership Climate" and Supervisory Behavior. Ohio 

State University, Personnel Research Board, 1951. 
The Foreman A Study of Supervision in British Industry. Staples 

Press, 1951. 
Foreman Training (Survey No. 8 of BNA's Personnel Policies Forum). 

Washington: Bureau of National Affairs, 1952. 

B. B. Gardner, Human Relations in Industry. Richard D. Irwin, Inc., 


C. S. Golden and H. }. Ruttenberg, The Dynamics of Industrial De- 

mocracy. Harper and Bros., 1942. 

W. A. Gomberg, Trade Union Analysis of Time Study. Science Re- 
search Associates, 1948. 

E. L. Hartley and R. E. Hartley, Fundamentals of Social Psychology. 
Alfred A. Knopf, 1952. 

G. W. Hartmann and T. Newcomb (Eds.), Industrial Conflict. The 
Cordon Co., 1939. 

W. R. Hayward and G. W, Johnson, The Story of Mans Work. Mil- 
ton, Balch and Co., 1925. (Quoted by courtesy of G. P. Putnam & 

H. Helson (Ed.), Theoretical Foundations of Psychology. Copyright 
1951, D. Van Nostrand Co., Inc. 

A. R. Heron, Why Men Work. Stanford University Press, 1948. 


L. H. Hill, Pattern for Good Labor Relations. Copyright 1947, 
McGraw-Hill Book Co., Inc. 

J. D. Houser, What the Employer Thinks. Harvard University Press, 

I. Howe and B. Widick, UAW and Walter Reuther. Random House, 
Inc., 1949. 

Interpreting the Labor Movement (Publication No. 9). Industrial Re- 
lations Research Assn., 1952. 

D. Katz, N. Maccoby, G. Gurin, and L. G. Floor, Productivity, Super- 
vision and Morale among Railroad Workers. University of Michi- 
gan: Survey Research Center, 1951. 

D. Katz, N. Maccoby, and N. C. Morse, Productivity, Supervision and 
Morale in an Office Situation. University of Michigan: Survey Re- 
search Center, 1950. 

F. P. Kilpa trick (Ed.), Human Behavior from the Transactional Point 
of View. Hanover, N. H.: Institute for Associated Research, 1952. 

A. Kornhauser (Ed.), Psychology of Labor-Management Relations. 
Champaign, 111.: Industrial Relations Research Assn., 1949. 

D. Krech and R. S. Crutchfield, Theory and Problems of Social Psy- 

chology. Copyright 1948, McGraw-Hill Book Co., Inc. 
K. Lewin, Dynamic Theory QJ; Personality. Copyright 1935, McGraw- 
Hill Book Co., Inc. 
K. Lewin, Principles of Topological Psychology. Copyright 1936, 

McGraw-Hill Book Co., Inc. 
J. F. Lincoln, Incentive Management. Cleveland: The Lincoln Electric 

Co, 1951. 
R. Linton, Acculturation in Seven American Indian Tribes. Copyright 

1940, Appleton-Century-Crofts, Inc. 
R. Linton, The Cultural Background of Personality. Copyright 1945, 

Appleton-Century-Crofts, Inc. 
N. R. F. Maier, Frustration. Copyright 1949, McGraw-Hill Book Co, 


N. R. F. Maier, Psychology in Industry. Houghton Mifflin Co, 1946. 
S. B. Mathewson, Restriction of Output among Unorganized Workers. 

The Viking Press, 1931. 

E. Mayo, The Human Problems of an Industrial Civilization. Harvard 

Business School, Division of Research, 1946. (First printing, New 

York: The Macmillan Company, 1933). 
E. Mayo, The Social Problems of an Industrial Civilization. Harvard 

University Press, 1945. 
E. Mayo and G. F. F. Lombard, Teamwork and Labor Turnover in the 

Aircraft Industry of Southern California, (Business Research 


Studies No. 32). Harvard University Graduate School of Business, 

Division of Research, 1944. 
W. McDougall, Introduction to Social Psychology. Methuen & Co., 

Ltd., 1908. 
H. C. Metcalf (Ed.), The Psychological Foundations of Management. 

Chicago: A. W. Shaw Co., 1927. 
}. G. Miller (Ed.), Experiments in Social Process. Copyright 1950, 

McGraw-Hill Book Co., Inc. 
J. Mills, The Engineer in Society. Copyright 1946, D. Van Nostrand 

Co., Inc. 

H. E. Mundel, Motion and Time Study. Prentice-Hall, Inc., 1950. 
G. Murphy, L. B. Murphy, and T. M. Newcomb, Experimental Social 

Psychology. Harper and Bros., 1937. 

National Institute of Industrial Psychology, Joint Consultation in Brit- 
ish Industry. Staples Press, 1952. 

T. M. Newcomb, Social Psychology. The Dryden Press, 1950. 
T. M. Newcomb and E. L. Hartley (Eds.), Readings in Social Psychol- 
ogy. Henry Holt and Co., 1947. 
C. Parker, The Casual Laborer and Other Essays. Harcourt, Brace, and 

Howe, 1920. 
J. M. Pfiffner, The Supervision of Personnel Human Relations in the 

Management of Men. Prentice-Hall, Inc., 1951. 

R. Presgrave, The Dynamics of Time Study. Copyright 1945, McGraw- 
Hill Book Co, Inc. 
A Review of Wage Incentive Practice. Melbourne, Australia: Industrial 

Welfare Division, Department of Labour and National Service, 

Commonwealth of Australia, 1949. 
F. J. Roethlisberger, Management and Morale, Harvard University 

Press, 1939. 
F. J. Roethlisberger and W. J. Dickson, Management and the Worker. 

Harvard University Press, 1939. 
J. H. Rohrer and M. Sherif (Eds.), Social Psychology at the Crossroads. 

Harper and Bros., 1951. 
A. M. Rose, Union Solidarity The Internal Cohesion of a Labor 

Union. Copyright 1952, University of Minnesota Press. 
F. H. Sanford, The Use of a Projective Device in Attitude Surveying. 

(Report No. 5). Haverford College and the Institute for Research 

in Human Relations, 1950. 
F. H. Sanford and I. M. Rosenstock, Projective Techniques on the 

Doorstep. (Report No. 7). Haverford College and Institute for 

Research in Human Relations, 1950. 


W. H. Scott, Industrial Leadership and Consultation. The University 
Press of Liverpool, 1952. 

M. Sherif, An Outline of Social Psychology. Harper and Bros., 1948. 

M. Sherif and H, Cantril, The Psychology of Ego-Involvements. Copy- 
right 1947, John Wiley and Sons, Inc. 

R. G. Smith and R. J. Westen, Studies of Morale Methodology and 
Criteria (Research Bulletin 51-29). San Antonio: USAF Air 
Training Command, Human Resources Research Center, 1951. 

E. Snow, People On Our Side. Random House, Inc., 1944. 

W. F. Taylor, The Principles of Scientific Management. Harper and 
Bros., 1911. 

L. R. Tripp (Ed.), Industrial Productivity, Madison, Wis.: Industrial 
Relations Research Assn., 1951. 

C. R. Walker and R. H. Guest, The Man on the Assembly Line. Har- 

vard University Press, 1952. 

H. F. Ward, In Place of Profit. Chas. Scribner's Sons, 1933. 
H. C. Warren, Dictionary of Psychology. Houghton Mifflin Co., 1934. 
G. Watson (Ed.), Civilian Morale, Houghton Mifflin Co., 1942. 
W. Williams, Mainsprings of Men. Chas. Scribner's Sons, 1925. 

D. Yoder, Personnel Management and Industrial Relations. Prentice- 

Hall, Inc., 1947. 

Youth and the World of Work. Michigan State College, Social Re- 
search Service, 1949. 

G. Zahnizer, G. Brown, R. D. Calkins, C. F. Mugridge, and C. Golden, 
Personnel Management Policies. New Wilmington, Pa.: The Eco- 
nomic and Business Foundation, 1944. 

Part I 


Industry's Core Problem in the 
Utilization oF Manpower 

THE EFFECTIVENESS of every business organization de- 
pends, in part, upon the perfection of the technical processes and equip- 
ment employed by it. Of at least equal importance in a company's success 
is the full utilization of its human resources. Of particular significance 
in this connection is the co-ordination of effort on the part of all employed 
in the organization from the office boy to the chief executive to ac- 
complish the purposes of the business enterprise. As Hayward and John- 
son, the authors of The Story of Man's Work, point out 

"There is one principle that has held good among workers ever since 
the day, far back when savage men still dwelt in caves, when for the 
first time two of them joined efforts to move a stone that was too heavy 
for one to lift. That is the principle that the work of the world goes 
forward most easily and swiftly when, and only when, all hands co- 
operate earnestly and willingly. That held good when the two hairy ape- 
men moved the stone. It held good when all the people, noblemen, yeo- 
men and serfs, joined to raise the medieval cathedrals. It holds good 
today, when in some giant corporation, with scores of factories and hun- 
dreds of thousands of workmen, all those workmen, all the managers, all 
the directors, and all the bankers who finance them, combine to turn 
out some product of every-day use so cheaply that the poorest man in 
America can have it in his home" (23, p. 22) . 


A common complaint in modern industry is that the drive towards 
such co-ordination of effort is lacking, particularly at the level of the 
rank-and-file employee. This shortcoming, according to many, reveals 
itself most prominently in the failure of workers to make full use of their 



capacities in maintaining and raising production standards. Strong, for 
example, has reached the conclusion that ''few, if any, employees are 
working up to their capacity" (44) . Stated in 1934, this conclusion might 
easily be regarded as descriptive of employee attitudes which arise only 
during a depression period. However, studies made during World War II 
and postwar years show that the problem of enlisting worker co-operation 
in a program of full production exists also during "prosperity" years ( 3, 
48). Approximately 55 per cent of executives of all types and sizes of 
manufacturing companies surveyed by Mill and Factory, in 1946, reported 
a decrease in labor productivity in their plants following World War II 
(28). As shown in Figure 1, a "general indifference on the part of labor" 
is referred to most frequently as the factor which accounts for the de- 
crease in employee efficiency. 

Production limits set openly by the union ^T 

Production limits sot secretly by the union 
Production limits set by the employes themselves 

A general indifference on the part of labor 

A "d^as-littU-as-nectssary" attitude on the part " 
of labor 

Inability of labor to produce 
at prewar level, due to: 


Poor supervision 

Lock of materials 
and parts 

Figure L Reasons Given by Executives for a Post- War Decline in Labor 
Productivity * (Note: These percentage figures do not add to 100% because 
many respondents checked two or more factors.) 

Difficulties experienced by individual firms in tapping the "will-to- 
work" are illustrated in a report published in 1946 by the National Indus- 
trial Conference Board (34), in which Henry Ford II is quoted as saying 
that the Ford Motor Company was getting about 34 per cent less output 
than it did in 1941, per direct labor man-hour, on comparable products 
and processes. The experience of the General Motors Corporation shows 
that, on operations which provided unquestionable bases for direct com- 
parisons, a thousand man-hours of direct labor produced only 63 per cent 
as much finished material in 1945 as in 1941, a 37-per-cent decline in 


* By permission, from "Labor productivity questions on the productivity of labor," 
til and Factory, (August) 1946. 


production which is close to the figure cited by the Ford Motor Company. 
An investigation made by the General Motors Corporation to determine 
the causes of losses in production led to the conclusion that approximately 
43 per cent of the total decline in production represented losses that could 
be laid at the door of management (resulting from poor organization 
and tooling, less experienced supervision and setup men, interruptions 
in the flow of materials, etc.). The remaining approximately 57 per cent 
of the total decline was due, in management's opinion, to reduced pro- 
ductive worker effort (9). 

As suggested above, many factors operate in raising or lowering pro- 
ductivity 1 (18, 63). An analysis of work standards by Wrape, in 1952, 
leads to the conclusions that in industries where man-hour output is rising, 
factors other than worker performance account largely for the productivity 
increase, and that "individual workers may actually be producing at a 
lower rate" than in earlier years (62, p. 64) . Growing demands by unions 
for increased wages keyed to a "productivity" or "annual improvement 
factor" reflecting increases in the nation's industrial output (37, 45a) 
bring into relief the question of the extent to which such increases reflect 
increased worker co-operation or effort. This question has been considered 
in a survey of opinions bearing upon employee contribution to produc- 
tivity. This involved 156 companies, representing a cross section of in- 
dustry and employing 1,500,000 workers. Replies from these companies; 
rank "increased labor efficiency" last in a list of seven factors believed to 
"contribute most" to total company productivity (32). In fact, the posi- 
tion taken by most of the companies is that in asking for productivity pay 
raises, labor "is asking for recompense for something that it had little or 
no part in achieving." 

The influence of employee motivation in accounting for low produc- 
tivity, and resulting increased unit labor costs, is illustrated in the findings 
of the 1945 nation-wide survey conducted for Factory by the Opinion 
Research Corporation. This showed that only 49 per cent of all manual 
workers, and an even smaller percentage of union manual workers think 

1 The terms production and productivity are used frequently throughout this book. 
Diebold defines productivity as "the quotient obtained by dividing production by 
one of the factors of production," thus making it "possible to speak of the productivity 
of capital, of raw materials, and of any other factor desired.'* The same authority 
notes that "productivity is most commonly expressed in terms of human labor" (18, 
p. 53). It is in this sense that productivity is viewed as a ratio of output to labor 
hours, and expressed in terms of output per man-hour. There is, however, con- 
siderable lack of both uniformity and clarity in the use of the term and, as Diebold 
has pointed out, "the word 'productivity' has come to be something of a five-syllable 
synonym for output" (18, p. 53). In general, the present author has avoided the 
discussion of this semantic problem, and has followed the usage of investigators cited 
in this text in referring to their research or views on production and productivity. 


that "when a man takes a job in a factory he should turn out as much 
work as he can' 9 (61). Another example is found in a survey conducted 
in 1951 by the Opinion Research Corporation, involving a national 
sample of 1,245 manual and white-collar employees of manufacturing 
companies (15). When asked "Do you feel the workers in your depart- 
ment (office) are already turning out as much work in a day as they 
can, or could they do more?" 34 per cent replied "Could do more" and 
another 11 per cent "Some could; some couldn't/' Only about one half 
of those interviewed ( 51 per cent) were of the opinion that they and their 
fellow workers were "turning out all we can! 9 

The Economic and Social Significance of High 


Such facts suggest that among the continuing needs of modern industry 
is a better understanding of the factors which underlie the will-to-work 
and of the conditions or devices which can most effectively arouse or 
release these inner forces which lead the employee to participate willingly, 
fully, and satisfyingly in furthering the production aims of industry. 


The advantages to the industrial enterprise of any program which will 
lead to increased production need no lengthy discussion. Differences in 
production among workers which have significance from the viewpoints 
of both employee earnings and company profits (18) are illustrated in a 
comparison of the productivity of workers on identical tasks in two British 
factories, reported by Marriott (33), Although widely separated and un- 
der different local managers, the two factories, designated as Factory X 
and Factory Y, were similarly organized and had the same higher manage- 
ment. In Figure 2 are shown productivity curves for two years ( 1947-48), 
based on four-week averages, of the number of man-hours taken to com- 
plete similar units of work. Analysis of data for the entire two-year period 
showed that workers in Factory X had, on the average, taken 178 man- 
hours and in Factory Y 255 man-hours, or 43 per cent longer, to complete 
a similar unit of work. On the other hand, as is apparent from Figure 2, 
there were times at the beginning and end of the two-year period when 
productivity was almost equal in the two plants. 

In an interim report it was stated by Wyatt (who conducted the study 
described by Marriott) that "the differences in output were due primarily 
to human factors and not to material differences in type or conditions 
of work." A further analysis of some of the factors which appeared to 






Figure 2. Differences in Productivity on Similar Work.* Productivity is 
plotted in terms of number of man-hours taken to complete similar units of 

.... "Standard Times" (Factory X) 

"Standard Times" (Factory Y) 

account for the observed difference in output has been presented by 
Marriott, as follows 

1. Mistakes made by the time-study engineers contributed to the 
difference. Workers in both factories were paid a bonus on the time 
saved on the "standard time/' Time studies were made too hurriedly, 
and standards were set so that workers in Factory Y had to do 20 per 
cent more work to reach the bonus level than those in Factory X. In 
fact, because of discontent and low output, increases of 44 per cent in 
"standard times" had to be made as well as several temporary allowances 
granted in Factory Y before agreement was reached. By contrast, in Fac- 
tory X, because of increasing efficiency, "standard times" were reduced 
by 6 per cent over the two-year period. 

2. Employees in Factory Y were, on the average, 6% years older than 
in Factory X; much less fit for strenuous work; not so carefully trained, 
and two thirds as compared with one third in Factory X had over a one- 
hour journey to and from work, which required much earlier rising. 

3. As reported by managers and supervisors, workers in Factory Y 
were either ex-miners or came from mining families. Their attitude to 
work apparently reflected "their past experience of hard times in a de- 

* By permission, from R. Marriott, "Socio-psychological factors in productivity," 
Occup. PsychoL, 1951, 25, 15-24. 


pressed area." They were suspicious and mistrustful, and efforts to win 
their confidence and co-operation apparently merely served to increase 
their mistrust (33, p. 23). 

The above example makes clear that employee productivity is the end 
effect of many influences. Although not noted by Marriott, the marked 
rise in production coincident with the increase in "standard time" in 
Factory Y may well illustrate the influence of the wage incentive itself, 
when based on standards which are carefully developed and viewed as 
"fair" by the worker. Suitability for the job; the nature of the training 
program; the pattern of the task, and many "physical conditions" such 
as hours of work, temperature, lighting, construction and speed of ma- 
chines, etc., are among factors underlying employee performance (50). 
However, there is also considerable evidence that the potential benefits 
of good selection, a well developed training program, good work methods, 
appropriate work standards and wage incentives, etc., can be considerably 
reduced or even nullified where, for one reason or another including a 
suspicion or distrust of management employees are unwilling or disin- 
clined to use their capacities or other media at their disposal for achiev- 
ing maximum output and maintaining high production standards. 


Maintenance of high production is of interest not alone from the view- 
point of the single plant, but also in terms of achieving broader economic 
and social objectives (45, 46). One basic test of any civilization is the 
extent to which it improves the material conditions of those who live 
under it (47). From this viewpoint, the effectiveness of the individual 
business enterprise and of an economic system as a whole can be measured 
by the availability, in the market place, of goods which minister to the 
satisfaction of human wants. Raising the level of consumption and 
standard of living involves lowering prices and increasing wages as a basis 
for the wider distribution of goods and of services. Such advances cannot 
be accomplished if output is restricted by employees or if other influences 
operate to raise unit output cost. Deficiencies in the motivation of em- 
ployees can be an important factor in such variations in production. Deal- 
ing with the problems and sources of employee motivation represents a 
potent tool for achieving necessary industrial and social objectives in the 
way of increased productivity. 

The Need for Job Satisfaction 

The excellence of a civilization is to be gauged not alone by a material 
yardstick, but also by the opportunities which it provides for the intellec- 


tual and emotional expression and development of the individual. There 
is little merit in a civilization which dulls the mind, warps the emotions, 
destroys the will, and reduces the individual to an automaton, even though 
it succeeds in providing an ever increasing supply of material goods for 
general distribution (49) . For this reason, it is not possible to think of the 
human element in industry in terms of productive efficiency alone. The 
psychologist must take full account of the possible consequences of our 
industrial order in mechanizing the mind, creating mental conflicts, 
diminishing creative power, and setting the stage for individual dissatis- 
faction and maladjustment at work. In other words, the psychologist (and 
also other social scientists) must be concerned with the satisfactions de- 
rived by the individual from the job. 

A common charge against modern industry is that it has made work 
dull and spiritless. Through the coming of the Machine Age, it is said, the 
workman has lost his joy in production (4, 59). "Now," says Morgan, 
". . . work is so specialized, so devoid of intrinsic interest that the work- 
man finds no incentive to work . . . The nature of the daily work of 
most of the working people precludes the possibility of their loving the 
work. Most of them hate it, and how can they help hating a job which 
means, for instance, that they go through a set of motions (which they 
learned in a very short time) hundreds of times a day with the prospect of 
day after day, week after week, year in and year out doing the same thing?" 
(35, p. 207). 


As this author has shown elsewhere (50, 52, 53), both a review of the 
history of man's work through the ages and modern studies of repetitive 
work raise questions concerning the validity of such an extreme indict- 
ment of the Machine Age. Nevertheless, there is considerable evidence 
that many workers are dissatisfied with their jobs. For example, according 
to a survey conducted for Fortune by Roper (40), in 1947, 20 per cent 
of a nation-wide sampling of factory workers found their jobs to be dull 
and monotonous most or all of the time (Table 1). In a study conducted 
by Centers (11), involving a sampling of the entire occupational stratifica- 
tion of the United States, discussed in some detail in later chapters, 
13 per cent of all manual workers were found to be dissatisfied with their 
jobs. A survey involving a sampling (N = 976) of men in two mass- 
production factories and six metal-rolling mills in Great Britain showed 
that from 59 to 75 per cent of the men interviewed expressed themselves 
as "satisfied" or "very satisfied" with the operations they performed ( 56) . 
Studies by Hoppock (24), Bell (6), Flowers (16), Habbe (21), Form 
(17), and others, summarized through the years by Hoppock and his 



associates in Occupations, suggest that 20 to 25 per cent represents a fair 
estimate of the proportion of dissatisfied workers, and that the percentage 
of workers dissatisfied with their jobs is particularly high at the lower levels 
of the occupational hierarchy. Similar findings are reported from surveys 
of workers in Holland (42). 

TABLE 1. Summary of Replies on Attitudes toward Present Job Fortune 
Survey of Public Opinion, (May) 1947 (after Roper) 

"Which one of these statements comes closest to describing how you feel about 
your present job?" 

Per cent of each group replying 








1. My job is interesting 

nearly all the time. 43 48 50 47 

2. While my job is interesting 
most of the time, there are 
some dull stretches now 

and then. 34 36 28 35 

3. There are a few times when 


my job is interesting but 
most of it is pretty dull and 

* ."* 







4. My job is completely dull 

and monotonous; there is 

nothing interesting about it. 






5. Expresses no opinion. 







According to Watson, "an economist, gifted with both competence and 
imagination, once defined his Utopia as a world in which everyone would 
be eager to do the kind of work he was doing, whether he was paid for it 
or not" (58, p. 114). There is, of course, a serious question as to whether 
this ideal can ever be realized. However, there also seems no question that 
much can be accomplished in increasing the extent of intrinsic job satis- 
faction, i.e., the satisfaction that derives from the content of the job it- 
self (27), through the application of vocational guidance and selection 
techniques; by reorganizing the job itself, e.g., through job enlargement 
and job rotation in repetitive jobs (38a, 54); by eliminating "deperson- 
alization" through the re-creation of a "bona fide work community" (55), 
and through other media discussed in texts on industrial psychology and 
industrial sociology. Although Utopia is a far-distant creation of man's 


imagination, it seems entirely possible to accept the suggestions, by Bal- 
chin, that work "should [and can] be something infinitely more pleasur- 
able and more satisfying a pleasant reality rather than a mere escape 
from reality" and that "an effort should be made to give the average man 
what many of us already enjoy a job which is not merely a means of 
'earning a living' but which has in it all the ingredients of life itself (4) . 
There is no question that the accomplishment of such an objective 
is extraordinarily difficult. However, there is also no question that increas- 
ing intrinsic job satisfaction through the modification of work proc- 
esses and the use of other devices represents a worthwhile goal in indus- 
trial management. There is a need for executives to recognize that "those 
who would be horrified at the suggestion that they spend their working 
lives doing something completely dull and pointless have little right to 
complain if other men now demand that they should enjoy their working 
lives too," and that, furthermore, "the demand for an enjoyable working 
life will be made ... is being made now, though in a muddled and in- 
coherent way" (4). Other considerations point to the desirability of tak- 
ing steps to reduce the incidence of dissatisfaction with the job. One of 
these is the practical possibility that workers with higher intrinsic job 
satisfaction are the more productive workers, at least in some job situa- 
tions. A few studies have, it is true, shown that high productivity is not 
necessarily a function of job satisfaction (27, 60). Other studies, such as 
that by Katz and Hyman (26), have revealed a positive relationship be- 
tween job satisfaction and production. Whether or not satisfaction is 
related to production, the need for steps to promote the feeling "I am 
satisfied vnth my job" (24) is supported by findings, discussed in later 
chapters, which indicate that job satisfaction can be and is an important 
ingredient of employee morale. 

Employee Morale 

Both everyday observation and studies described in this volume indicate 
that independent organizations and departments within the same organ- 
ization differ widely with respect to levels of employee morale. The term 
morale has been used loosely, both by industrial people and by psycholo- 
gists. In many instances, as is evident from studies described in Part III 
of this volume, the term morale has been used as equivalent to intrinsic 
job satisfaction. However, it has become increasingly evident, as Katz 
points out, that "morale is not a strictly unitary concept, but that it con- 
sists of a number of dimensions" (27) . There is no complete agreement as 
to what these are. In general, particularly in earlier studies, the trend has 
been to define significant dimensions in terms of what is revealed in 
individual experimental studies and attitude surveys. More recently, there 


has been a growing inclination to define morale primarily in terms of atti- 
tudes and behavior traits associated with or derived from the activity of 
the individual as a member of a group. In addition, emphasis has been 
placed upon the implications of morale for individual adjustment, par- 
ticularly in terms of the orientation of the individual toward future goals 
toward "effective and confident striving for conditions that will work 
for good adjustment in the future" ( 12, p. 394) . 


In spite of differences of opinion, 2 discussed further in Chapter 14, there 
is considerable support for the position that 

"Morale is an attitude of satisfaction with, desire to continue in, 
and willingness to strive for the goals of a particular group or or- 
ganization" (41, p. 1). 

An important element in high morale is the feeling on the part of the 
individual that "he shares the basic purposes of the groups of which he 
is a member" a state of mind which ''makes it possible for him to per- 
form his tasks with energy, enthusiasm, and self-discipline, sustained by a 
conviction that, in spite of obstacles and conflict, his personal and social 
ideals are worth pursuing" (12, p. 394). From the viewpoint of manage- 
ment, the problem of employee morale is that of stimulating a feeling of 
"togetherness," of "a sense of identification with an interest in the ele- 
ments of one's job, working conditions, fellow workers, superiors, em- 
ployers, and the company" (7, p. 299) conducive to the achievement of 
the organization's aims. High employee morale is important because 
plant productivity and efficiency of operation depend upon employee co- 
operation in attaining necessary output standards. High employee morale 
is equally important because of the implications of low morale for in- 
dustrial conflict. 


There are many indications of continuing hostility and strife in the 
industrial scene. A significant index of the extent of such conflict is found 
in the number of strikes. During the five-year period 1948-1952 there 
have been 21,468 work stoppages (caused almost entirely by strikes) in- 
volving approximately 13 million workers and resulting in the loss of 
about 201 million workdays. 

Cost of Strikes: The cost of strikes, to industry and to society at large, 
is illustrated in figures applying to the 1949-50 coal strike in the United 

2 See particularly Allport (2), Child (12), Lippitt (29), Maier (31), Watson 
(57), and Stouffer (43). 


States. According to a report of the National Association of Manufac- 
turers (14), based on figures collected by the U.S. Bureau of Labor Sta- 
tistics, 23,790,000 man-days were lost in the bituminous coal industry dur- 
ing the actual strike period, September 19, 1949, to March 5, 1950. 

Assuming an average daily pay of $15.46, the time lost because of 
strikes alone cost the miners approximately 370 million dollars. On the 
assumption that on the average 325,000 men were made idle during the 
strike period, these workers sustained a pay loss of about $1,200 each. In 
the final strike settlement, miners were awarded a pay increase of 70 
cents a day. The average worker, therefore, would have to work 1,714 days 
to make up the loss sustained because of the direct strike alone. Since the 
average workyear in the industry, as calculated for the period 1940-47, 
included 237 working days, the average miner would have to work a little 
over seven years to make up the pay lost because of the direct strike alone. 
Although this loss of wages is counterbalanced by other major monetary 
gains particularly the 10-cents-a-ton royalty for the miners' welfare fund 
the direct net wage loss to the individual miner still remains substantial. 

As is pointed out in the report under consideration, loss of purchasing 
power of the miners, amounting to at least 370 million dollars, not only 
imposed hardships on the workers involved, but had far-reaching effects 
on the entire economy. Furthermore, both the coal industry and the na- 
tional economy suffered from the great loss of coal production during the 
strike. It is pointed out that it is difficult to estimate with accuracy the 
size of this loss. However, consideration of figures given by the U.S. 
Bureau of Labor Statistics and the U.S. Department of Commerce leads 
to the estimate that, for both 1949 and the first two months of 1950, about 
221 million tons of bituminous coal production was lost. 

Strikes bring not only economic losses to all involved to workers, com- 
panies, and society alike but they represent an overwhelming accumula- 
tion of individual and group tensions which can have disastrous effects 
on the stability of the economic system. Throughout the world, except in 
the U.S.S.R. and other Communistic countries, where labor unions are 
dominated by government (19, 51) and strikes are not permitted (20), 
strikes have seriously prejudiced both the economy and the social order. 


Although strikes, with their terrific impact upon public opinion (1), 
represent the most evident sign of industrial strife, the latter expresses 
itself in many different situations and with many degrees of severity. In 
a large sense, every grievance in a plant is an expression of conflict between 
management and the workers. Conflict on a broader scale and of greater 
severity is found in repeated instances of both organized and unorganized 


resistance to technological and other changes. As McGregor and Knicker- 
bocker (30) have pointed out, even minor changes in work processes and 
equipment or in procedures may have major consequences for employee 
morale and industrial relations. But there can be no "right approach" 
in solving the problem of resistance to change and eliminating the source 
of conflict at its base without reference to "group thinking" ( 5 ) without 
reference to the sentiments and feelings of the employees involved and of 
their fellow workers (39, p. 580). 


The need for reducing industrial strife, as well as requirements in 
the way of enlarged productivity and job satisfaction, make it necessary 
to gather accurate information on "the biological and derived needs, the 
interests, the dissatisfactions, and the aspirations" of individuals working 
together in industry (10) . There is an urgent need for the delineation of 
the dimensions of morale and of the sources and determinants of em- 
ployee motivation. One purpose of this volume is to describe what has so 
far been accomplished in these directions. Another and major purpose 
is to indicate the implications of such facts (and of the methods by which 
they are derived) to management particularly in terms of management's 
function in the development and administration of employee-relations 
and labor-relations policies and practices. 

Basic throughout this volume is the conviction, so well stated by 
Drucker, that u the human being is the central, the rarest, the most pre- 
cious capital resource of an industrial society" 8 (13, p. 56). Basic also is 
the conviction that management has a fundamental responsibility as a 
"coordinator of human efforts" in industry. Management must meet this 
responsibility, not alone in the interest of the individual enterprise, but as 
a necessary step in preserving the right and opportunity for industry to 
operate as a private enterprise within a free and democratic society. 

Man has progressed from savagery to civilization largely by learning 
how to release, direct, and control the energies of inorganic matter. Prog- 
ress toward a better civilization now depends largely upon his achieve- 
ments in releasing, directing and controlling the energies of man himself. 
In no place is this more apparent than in man's daily task of producing 
and distributing those material goods upon which his civilization depends. 
The disregard of a worker's capacity to feel, think, and grow is a subtle but 
menacing danger in breaking down his social and spiritual morale (25). 
To increase productivity, heighten job satisfaction, and raise the level of 
employee morale, it is necessary to arouse the intelligent interest of the 
employee. It is urgent to enlist his feelings as well as his abilities in his 

* Not italicized in the original. 


work. The failure to do so will, at best, produce an ineffective worker. At 
worst, it will transform the worker into an industrial rebel. The indi- 
vidual worker's philosophy is built upon the concrete experiences which 
personally affect him. From these experiences, multiplied millions of 
times the country over, may easily emerge a cynical or intensely resentful 
group philosophy extremely dangerous to the democratic government and 
industrial organization of contemporary America. 


1. The Aftermath of Strikes. Princeton: Opinion Research Corporation, 1946. 

2. G. W. Allport, "The nature of democratic morale" in G. Watson (Ed.), Civilian 
Morale. Houghton Mifflin Co., 1942, 3-18. 

3. H. A. Balke and G. C. Thompson, "Productivity stands still/' Conference Board 
Bus.Rec., 1952,9,72-77. 

4. N. Balchin, "Satisfactions in work/' Occup. PsychoL, 1947, 21, 125-34. 

5. S. Barkin, "Human and social impact of technical changes/' Proceedings of Third 
Annual Meeting, Industrial Relations Research Assn. 1950. Champaign, 111.: 
Industrial Relations Research Assn., 1951, 112-27. 

6. H. M. Bell, Youth Tell Their Story. Washington: American Youth Commission. 

7. A. B. Blankenship, "Methods of measuring industrial morale" in G. W. Hart- 
mann and T. Newcomb (Eds.), Industrial Conflict. The Cordon Co., 1939, 299- 

8. A. H. Brayfield and F. H. Rothe, "An index of job satisfaction, " /. Appl. PsychoL, 
1951, 35, 307-11. 

9. S. L, H. Burk, Labor Relations and Competitive Costs Hindrances to Full 
Production. An address at the Industrial Relations Conference, Chamber of 
Commerce of Phila., Feb. 5, 1948, 6-8. 

10. H. Cantril and D. Katz, "The problem of method" in G. W. Hartmann and T. 
Newcomb (Eds.), Industrial Conflict, The Cordon Co., 1939, Chapter 1. 

11. R. Centers, "Motivational aspects of occupation stratification," J. Soc. PsychoL 
1948, 28 (Second Half), 187-217. 

12. I. L. Child, "Morale: a bibliographical review," PsychoL Bull., 1941, 38, 393-420. 

13. P. F. Drucker, "Basic elements of a free, dynamic society: Part 1," Harvard Bits. 
Rev., 1951, 29, 6, 56-57. 

14. The Economic Impact of an Industry-wide Strike. (Economics Policy Division 
Series No. 27). New York: National Association of Manufacturers, 1950. 

15. Employee Cooperation on Productivity. Princeton: Opinion Research Corpora- 
tion, 1951. 

16. M. Flowers, "Young America in revolution," World's Work, 1927, 54, 273-80. 

17. H. Form, 'Toward an occupational social psychology/' /. Soc. PsychoL, 1946, 24 
(First half), 85-99. 

18. J. A. Diebold, The significance of productivity data," Harvard Bus. Rev.. 1952, 

19. M. Gordon, Be/ore and After Lenin. E. P. Button & Co., Inc., 1941. 

20. V. Gsovski, Elements of Soviet Labor Law (Bureau of Labor Statistics No. 1026). 
U.S. Govt. Printing Office, 1951. 


21. S. Habbe, "Job attitudes of life insurance salesmen," /. Appl. Psychol, 1947, 31, 

22. J. Harding, "Measure of civilian morale" in H. Cantril (Ed.), Gauging Public 
Opinion. Princeton University Press, 1944, 233-58. 

23. W. R. Hayward and G. W. Johnson, The Story of Man's Work. Minton, Balch 
& Co., 1925. (Quoted by courtesy of G, P. Putnam Sons). 

24. R. Hoppock, Job Satisfaction. Harper & Brothers, 1935. 

25. J. D. Houser, What the Employer Thinks. Harvard University Press, 1927. 

26. D. Katz and H. Hyman, "Morale in the war industry" in T. Newcomb and E. 
Hartley (Eds.), Readings in Social Psychology. Henry Holt & Co., 1947, 437-47. 

27. D. Katz, "Morale and motivation in industry" in Current Trends in Industrial 
Psychology. University of Pittsburgh Press, 1949, 145-71. 

28. "Labor Productivity Questions pn the Productivity of Labor" (Mill and Fac- 
tory Survey of the Month), Mill and Factory, 1946 (August). 

29. R. Lippitt, "Morale of youth groups" in G. Watson (Ed.), Civilian Morale. 
Houghton Mifflin Co., 1942, 119-42. 

30. D. McGregor and I. Knickerbocker, "Industrial relations and national defense: 
a challenge to management," Personnel, 1941, 18, 1, 49-63. (Reprinted as 
M.I.T. Dept. of Economics and Social Science, Series 2, No. 7). 

31. N. Maier, Psychology in Industry. Houghton Mifflin & Co., 1946, 80-108. 

32. Management and People Newsletter, Fisher, Rudge and Neblett, Inc., May 27, 

33. R. Marriott, "Socio-psychological factors in productivity," Occup. Psychol, 1951, 
25, 15-24. 

34. Measuring Labor's Productivity ^ .(Studies in Business Policy No. 15), New York: 
National Industrial Conference Board, 1946. 

35. J. J. B. Morgan, "Why men strike," Am. J. Soc., 1920, 26, 207-11. 

36. R. B. Perry, 'The meaning of morale," Educ. Rec., 1941, 22, 446-60. 

37. The Productivity Factor in Wage Adjustments Selected References. Princeton 
University, Industrial Relations Section, 1952 (May). 

38. Productivity from the Worker's Standpoint. Princeton: Opinion Research Corpo- 
ration, 1946. 

38a. S. A. Raube, 'The problem of boredom," Conference Board Mgmt. Rec., 1948, 
10, 565-76. 

39. F. J. Roethlisberger and W. Dickson, Management and the Worker. Harvard 
University Press, 1939. 

40. E. Roper, "The Fortune Survey/' Fortune, 1947 (May), 5-12. 

41. R. G. Smith and R. J. Westen, Studies of Morale Methodology and Criteria 
(Research Bulletin 51-29). San Antonio: USAF Air Training Command, Human 
Resources Research Center, 1951. 

42. J. Stapel, "Living research," Pub. Opin. Quart., 1950, 14, 553-54. 

43. S. A. Stouffer et al, The American Soldier: Adjustment During Army Life 
(Vol. I. Studies in Social Psychology in World War II). Princeton University 
Press, 1949. 

44. E. V. Strong, "Aptitudes versus attitudes in vocational guidance," /. Appl. Psychol., 
1934, 18, 501-15. 

45. L. R. Tripp (Ed.), Industrial Productivity. Madison, Wis.: Industrial Relations 
Research Assn., 1951. 

45a. L. S. Stessin, "Productivity raises," Forbes, 1952, 70, 7, 31. 


46. Summary of Proceeding? of Conference on General Productivity (U.S. Dept. of 
Labor, Bureau of Labor Statistics, No. 913). U.S. Govt. Printing Office, 1946. 

47. H. H. Tullis, "Wage administration and incentives," Proceedings: Seventh Inter- 
national Management Congress, General Management Papers, The Williams and 
Wilkins Co., 1938, 29a-29d. 

48. A. Uris, "How good a leader are you ... at getting better production?/' 
Factory Mgmt. Mainten., 1951, 109, 7, 68-70. 

49. M. S. Viteles, "The role of industrial psychology in defending the future of 
America," Ann. Amer. Acad. Pol. Soc. Science, 1941 (July), 216, 156-62. 

50. M. S. Viteles, Industrial Psychology. W. W. Norton & Company, Inc., 1932. 

51. M. S. Viteles, "Industrial Psychology in U.S.S.R.," Occup. PsychoL, 1938, 12, 

52. M. S. Viteles, The Science of Work. W. W. Norton & Company, Inc., 1934 
Chapters 10 and 11. 

53. M. S. Viteles, "Man and Machine relationship; the problem of boredom," Pro- 
ceedings of 1950 Annual Fall Conference, Society for Advancement of Manage- 
ment. New York: Society for the Advancement of Management, 1951, 129-38. 

54. C. R. Walker, "The Problem of the repetitive job," Harvard Bus. Rev., 1950, 
28, 54-58. 

55. C. R. Walker and H. R. Guest, The Man on the Assembly Line, Harvard Uni- 
versity Press, 1952. (See also G. R. Walker and H. R. Guest, "The Man on the 
Assembly Line," Harvard Bus. Rev., 1952, 30, 71-83.) 

56. }. Walker and R. Marriott, "A Study of some attitudes to factory work," Occup. 
Psychol. London, 1951, 25, 181-191. 

57. G. Watson, "Morale during unemployment" in G. Watson (Ed.), Civilian 
Morale. Houghton Mifflin Co., 1942, 273-348. 

58. G. Watson, "Work satisfaction" in G. W. Hartmann and T. Newcomb (Eds.). 
Industrial Conflict. The Cordon Co., 1939, 114-24. 

59. S. Webb and B. Webb, "Labor" in C. A. Beard, Whither Mankind? Longmans, 
Green & Co., Inc., 1928, 140-41. 

60. I. Weschler, M. Kahane and R. Tannenbaum, "Job satisfaction, productivity 
and morale: a case study," Occup. Psychol., 1952, 26, 1-14. 

61. "What the factory worker really thinks about productivity, nationalization of 
industry, and labor in politics," Factory Mgmt. Mainten., 1946, 104, 1, 80-88. 

62. H. E. Wrape, "Tightening work standards," Harvard Bus. Rev., 1952, 30, 64-74. 

63. S. Wyatt, A Study of Variations in Output (Industrial Health Res. Bd. Emer- 
gency Report No. 5). H. M. Stationery Office, 1944. 

Wage-Incentive Systems 

WHY MAN works is a fundamental question in considering 
man's efficiency and adjustment in industry. Until recently, the answers 
to this question have been categorical and dogmatic in character. He 
works, according to the moralist, because of an inner urge which compels 
him to contribute his share as a member of the working society of men. 
Man works, according to the realist, because only by doing so can he 
feed himself and his family and provide for other wants. According to this 
view, he works because he is paid for it, and the amount of work he will 
do and the satisfaction he derives from it will depend upon the amount 
of pay he receives: in other words, upon the attractiveness of the monetary 

The application of the principles of scientific management in modern 
industry has, in fact, been accompanied by widespread dependence upon 
wage incentives to stir the worker to greater production ( 104) . In formu- 
lating the principles of scientific management, Taylor pointed to the need 
for "another type of scientific investigation . . . which should receive 
special attention, namely the accurate study of the motives which in- 
fluence man . . . Laws which apply to a large majority of men," he wrote, 
"unquestionably exist, and when clearly defined are of great value in deal- 
ing with men" (97). Nevertheless, in the day-to-day application of this 
system, increased pay for a higher rate or quality of work has constituted 
the major, if not the sole, incentive employed in elevating the worker 
to the higher standards of efficiency demanded by the system. Further- 
more, both management and union leaders have generally accepted the 
viewpoint that the desire for more money constitutes the prime, if not 
the sole, motive for increasing production, and that increased pay repre- 
sents the basic source of satisfaction in work. 1 Evidence of the prevalence 

i Sec Chapter 15. 


of this opinion is found, for example, in a survey reported in the Public 
Opinion Index for Industry in 1947 (73). Questioned as to what is of 
importance in gaining worker productivity, 50 manufacturing executives 
replied as follows 



"Money alone is the answer" 22 44 
"Money is by far the chief thing but some importance 

is to be attached to less tangible things.' 9 14 28 
"Money is important but beyond a certain point it will 

not produce results." 14 28 

A question, in the same survey, requesting opinions on the best wage 
system, showed almost unanimous agreement that some wage-incentive 
plan, such as straight piece work, bonus pay, etc., stimulates more pro- 
duction than straight hourly pay. 

Variety of Wage- Incentive Systems 


Emphasis upon earnings as a source of motivation has led to the devel- 
opment and use of a large variety of wage incentive plans. 2 Common 
among these are various kinds of piece-work systems under which pay- 
ment is in relation to the number of units produced by the worker (80). 
In straight piece work, payment is at a single flat rate per unit of output. 
Under a differential piece-rate plan, as first formulated in the 1880's by 
Taylor, each operation is assigned two piece prices, one from 30 per cent 
to 50 per cent higher than the other (41). The worker who completes 
in a day a number of pieces less than the standard task, as established by 
time studies, is paid at the lower rate. The worker who completes more 
than the number of pieces included in the standard task is paid at the 
higher rate. Modifications of this plan are to be found in wage-incentive 
systems developed by Gantt (25) and in the Multiple Piece Rate (41) 
plan, the latter being characterized by three distinct piece prices; a lower 
one for production up to 83 per cent of the standard task; a middle one 
applying to production between 83 per cent and 100 per cent, and a higher 
piece price paid for production above 100 per cent of the standard task. 

* For detailed description of wage incentive plans and their operation the reader 
is referred to Lytle (49), Balderston (2), Dickinson (17), Gomberg (31), Reitinger 
(77), Lincoln (44), Lynton (48), Russon (86), and to publication by the Amer- 
ican Management Association (40, 71), the National Industrial Conference Board 
(91, 109, 110), the Dartnell Corporation (23), the Industrial Relations Counselors, 
Inc. (18), National Association of Manufacturers (107), British Ministry of Labour 
and National Service (57) and the Australian Department of Labour and National 
Service (80). 



Under the task-bonus wage-incentive plan, also widely used in industry, 
standard job times, known as task times, are calculated, generally on the 
basis of time studies. The worker is paid at an established hourly rate and, 
in addition, receives a bonus if he performs the job in the standard time 
or does not exceed it by more than a stated percentage. Such a plan, as 
represented in the Gantt Task-Bonus Plan (25) and the Emerson Effi- 
ciency Plan (22, 77), the Bigelow Plan, and the Parkhurst Differential 
Bonus System (107) has the advantage of providing a guaranteed mini- 
mum wage for all employees, including inexperienced workers who cannot 
be expected to meet the production standards while learning the job. 

A further modification of such plans is to be found in premium-bonus 
systems, under which the worker does not receive extra payments unless 
he actually completes the job in less than the standard time, earnings con- 
sisting of time wages for the actual time taken to do the job plus payment 
for the time saved in comparison with the standard task time as estab- 
lished either on the basis of past experience or by time-and-motion-study 
techniques (80). Such plans vary with respect to the fraction of the time 
which is taken as the bonus for wage-incentive payment. Under the Halsey 
Plan (35), for example, the worker is paid on the basis of one-third of the 
time saved. 8 Thus, if the standard time is eight hours, and a worker com- 
pletes the job in six hours, he is paid at the established time rate for six 
hours, plus one-third of the two hours saved. In other words, he receives 
pay for 6% hours for the six hours actually taken to complete the job. 

Criticisms raised by workers against a plan which gives them only a 
portion of the savings effected by their efficiency has been overcome in the 
Standard Hour Plan, sometimes called a "100 per cent premium" plan, 
under which a standard work task, arrived at by time study, is expressed in 
standard hours, and the worker is paid the full savings resulting from 
production above standard. According to a report of the National Indus- 
trial Conference Board, an increasing number of American firms have 
adopted the Standard Hour Plan in recent years (110), 


A somewhat different approach in the development of wage-incentive 
plans is found in the point system, which evaluates a task in time units 
which supposedly take account of the amount of skill and energy required 
for its efficient performance and also of the amount of rest an average 

An interesting feature of the Halsey Plan is that it includes the payment of a 
wage incentive to supervisors, the payment made being fixed and paid from the em- 
ployer's share of the savings. 


worker needs during the performance of the task. The time unit used 
in the system is referred to as a "point" and each task is appraised as 
being worth a given number of points. Apart from the evaluation of the 
job in points, the wage-incentive principle is similar to that of the 
premium-bonus system in that the worker increases his earnings in propor- 
tion to the amount of time below the standard time or "bogey" in which 
he completes his task. 

The best known of the point plans is the Bedaux Point Premium Plan 
(80). The point in this system is called the "Bedaux" or "B," and is "a 
fraction of a minute of work, plus a fraction of a minute of rest, the aggre- 
gate of which is always one minute, but the proportions of which vary 
according to the nature of the operation" (9). The allowance for rest, 
usually known as the coefficient of rest, generally increases with the sever- 
ity of physical strain involved. 

Under the Bedaux system, a specific number of points is allotted to each 
of the tasks comprising a job. The amount of the bonus paid to the worker 
is determined, under this plan, by the number of B's in excess of 60, repre- 
senting a standard hour's work, which he accumulates by completing more 
than the required or "standard" amount of work. As this plan was origi- 
nally administered, the employee's bonus was 75 per cent of the value of 
the time represented by the points saved, the remaining 25 per cent being 
distributed among supervisory and managerial personnel in recognition 
of their share in making the saving which earned the bonus. 

In spite of certain advantages, especially its applicability to a large va- 
riety of jobs in the plant, the Bedaux system has been severely criticized 
because of the expense of installation and administration; because of 
difficulties in explaining it to workers, and especially, according to labor 
leaders, because the returns to the workers in the way of increased earnings 
are purportedly not commensurate with the increase in production under 
the plan as ordinarily administered (20). According to the 1949 survey 
made by the Division of Labour and National Service of the Common- 
wealth of Australia (80), few companies are now using the original plan, 
the general practice now being to pay employees the full value of the time 
saved, and the types of point systems in general use represent considerably 
modified adaptations of the original Bedaux Point Premium Plan. 


In addition to plans of the type described above, designed to provide 
wage incentives on the basis of individual increases in production, there 
are a number of group incentive plans (2) . These are intended primarily 
for use in situations where production represents the collective effort of 
employees whose work is related and interdependent. While group in- 


centive plans are primarily designed for use where the production of the 
individual cannot be directly evaluated, such plans have also been used, 
especially in Great Britain (56), in situations where individual incentive 
plans could be applied, largely because of their flexibility and their effect 
in promoting group or team spirit. According to a report of the National 
Industrial Conference Board, many firms in the United States adopted 
group plans in preference to individual plans during World War II, 
largely because they can be installed with less difficulty (110). 


Wage-incentive plans take many forms. Apparent in recent years is an 
increased effort to simplify the plans. One objective in such simplification 
is to overcome resistance by employees which arises from lack of under- 
standing of the plan and inability to perceive the direct relation between 
the effort and the reward. Illustrations of such an approach are to be 
found in the work of Scanlon (15, 88) in the United States and in 
that of Russon (86) in England. Plans developed under their auspices 
are generally alike in that they set production "targets" for a "target" date 
and use the difference between the norms (as established in the target) 
and actual performance as a base for bonus payments ( 50) . 

The first requirement in the application of the Scanlon Plan is to find 
a "normal" labor cost for the plant under consideration. Steps are then 
taken to develop means for giving labor the benefit of anything it can 
save under that "norm." The target is defined in terms of some tangible 
feature of the manufacturing situation. In a plant manufacturing silver- 
ware it is ounces of silver processed; in a wholesale warehouse it is tons 
warehoused; in a steel foundry and machine shop it is pounds of castings 
produced. In still another instance, in the case of a versatile steel fabricat- 
ing shop, it is a calculated percentage of operating profits per month. 

The system at the Lapointe Machine Tool Co. provides an illustration 
of how "norms" are set in the operation of a Scanlon plan, and how the 
amount available for distribution as a bonus to workers is determined. 
Here, where measurement was relatively easy, an "understandable ac- 
counting handle" the ratio of labor costs to total production value 
(monthly sales plus or minus the change in inventory) was used in 
setting the "norms." According to Davenport, Department of Commerce 
figures for 1947 showed that the ratio of wages and salaries to the value 
of shipments for the machine-tool industry throughout the country was 
roughly 41 per cent. Because of the feeling, on the part of the company, 
that the "norm" derived from its war record was too high, the "norm" at 
the Lapointe Machine Tool Co. was set at 38 per cent. Assuming the 
value of total shipments for a given month as $70,000, and an inventory 


change of plus $30,000, total production for that month would be 
$100,000. The anticipated or "normal" payroll for the month would be 
calculated at $38,000 ( 38 per cent of total production value) . If the actual 
payroll were only $35,000, a total of $3,000 would be distributed to the 
workers as a bonus (15). 

As stated earlier, a significant feature of the Scanlon Plan is that labor 
gets the entire amount in the way of savings in labor cost. "Management's 
profit from the plan," according to Davenport, "is derived from increased 
sales with no corresponding increase in total 'burden/ (i.e. overhead and 
labor costs) . . . Thus the basic theory is that labor should profit from 
extra earnings, while the company profits from a better use of its assets" 
(15). Another basic feature of the Scanlon Plan is that everybody in the 
plant participates in the bonus. 

The roots of what is generally known as the Scanlon Plan are found in 
union-management co-operation to save a marginal steel company from 
"going on the rocks" in the late 1930's. According to Daigneault, who has 
reviewed the operation of the Scanlon Plan in the plant of the Lapointe 
Machine Co., one advantage of the Scanlon Plan is that it actually serves 
to strengthen the union in a plant, as evidenced in the experience of that 
company over a period of five years (14a). The concept of union- 
management co-operation in the formulation of productivity plans ( 30, 
117) has remained basic in the installation of the Scanlon Plan, except 
in a few instances. As a result, independent evaluation of the effect of 
union participation both within and outside of collective bargaining 
agreements would be required in the appraisal of the Scanlon Plan in 
comparison with other wage-incentive plans. This basic question, which 
arises with increasing frequency in the operation of all kinds of wage- 
incentive systems, is discussed in later chapters, particularly Chapter 23. 


Profit-sharing plans, which provide for the distribution of a share of 
company profits, either in the form of cash or of company stock, have been 
viewed as an incentive to production. 4 In addition, such plans are gen- 
erally perceived as bringing significant by-products in the way of higher 
morale, increased "loyalty" and greater interest in the welfare of the 

The unfortunate repercussion from such plans during the 1930's de- 
pression years, and the decline in the number of such plans in 1929 have 
been cited as evidence of the limitations of profit sharing (24, 118) . Fur- 

* For descriptions of profit-sharing plans and their operation the reader is referred 
to publications by Balderston (3), Thompson (98), Rowe (84), Corby (14), and 
others (42, 114, 74, 76, 84, 69). /V ' 


thermore, it has been suggested that profit-sharing (including stock- 
sharing) plans appear to be weak as incentives to immediate production 
because of the "time lag in payment, the lack of worker responsibility for 
results, and the large size of the groups involved ... It is fairly certain/' 
according to Marriott, "that more distant goals (as represented in such 
plans) have comparatively little influence as compared with immediate 
needs" (53, p. 16). 

Prevalence of Wage- Incentive Plans 

Not only are there many different wage-incentive systems, but these 
have been widely used both in the United States and in other countries. 
Wage incentives have been applied not alone to raise production in fac- 
tories and offices, but also to increase sales and number of deliveries; to 
encourage more economical use of gasoline in the operation of motor 
vehicles; to improve the performance of supervisory and management 
personnel (76); to stimulate heightened achievement of personnel in 
almost every phase and at every level of manufacturing and commerce. 


In a survey conducted by the National Industrial Conference Board 
(91) in 1939, it was found that of 2,700 concerns in all types of business, 
employing about five million persons, 51.7 per cent, representing 53.3 
per cent of the total employees covered, were using wage-incentive plans 
of some kind in the United States. 

The demands for production during World War II, as well as the op- 
portunity to use incentive plans as a means of increasing wages despite 
wage stabilization (66), resulted in a renewed drive for the installation 
of wage-incentive plans supported, to some extent, by government officials 
and labor organizations (103). According to a report appearing in For- 
tune in 1943, such union support reflected the willingness of labor or- 
ganizations to accept wage-incentive plans as a way of circumventing the 
Little Steel formula, representative of the operation of "wage freeze" 
restrictions, as well as "an upsurge of interest in the incentive idea as a 
means of pushing out more material in the face of a declining labor 
supply" (108). Government support of financial incentives as a means 
of increasing productivity during the war is evidenced in the establish- 
ment, in 1945, of an Incentive Division in the Office of Commerce, U.S. 
Department of Commerce (66). 

In spite of the developments discussed above, neither government nor 
other surveys reveal an upsurge in the use of wage-incentive plans in the 
United States during or subsequent to World War II. Data which are 
most nearly comparable to those cited for earlier years are found in a 


study made late in 1946 by the National Industrial Conference Board 
(67), involving 3,498 companies employing approximately 6% million 
persons. 6 According to this report, incentive payments for "productive" 
work decreased from 517 per cent in 1939 to 49.7 per cent in 1946. As in 
1939, a separate study was made of selected companies (300 in number) 
which offered "an even choice between day work and incentive systems/' 
Approximately 77 per cent of such companies were found to be using some 
kind of financial incentive as compared with 75 per cent in 1939. 

In contrast to the figures cited above are data from government surveys 
(102), made in 1945 and 1946, which show that only about 30 per cent 
of manual workers in manufacturing industries (omitting basic iron and 
steel, printing, rubber and lumber) were receiving incentive payments. 
While the data from these independent surveys differ, in part because of 
differences in coverage, the facts appear to support the conclusion, stated 
above, concerning the absence of an upsurge in the use of incentive sys- 
tems in the United States during the decade 1940-50, and suggest that 
there may actually have been some reductions in use, in spite of the forces 
operating in favor of the installation of wage-incentive plans during the 
war years. Nevertheless, according to a report by Ramond, in 1947 a total 
of 7 out of 12 million industrial workers in the United States were being 
paid according to some form of wage incentive system (113). 


As of 1947, 26 per cent of manual workers in Great Britain were being 
paid piece rates or receiving some kind of bonus, while 74 per cent were 
paid on a time-rate basis (55). A survey, in 1949, of 25 per cent of em- 
ployees in private firms subject to payroll tax in Australia showed that 
approximately 28 per cent were being paid on the basis of some kind of 
wage-incentive plan, the figure for manual workers being approximately 
32 per cent (39). In both of these countries, even under labor govern- 
ments interested in the establishment of socialistic states, there has been 
wide and continuing interest in the possibility of stimulating production 
through the wider use of wage-incentive plans. As in the United States f 
there has been considerable controversy concerning the merit of various 
plans, e.g. individual versus group plans. In Great Britain where, accord- 
ing to Marriott, "incentives are now (1951) almost as topical as the 
weather" (53), reports by teams organized under the Anglo-American 
Council on Productivity to survey methods and particular industries, e.g. 

The incentive-wage plans considered did not include profit-sharing and employee- 
suggestion systems. 

6 Data on "nonproductive" workers were collected in the 1946 but not in the 1939 


petroleum (79) and building (78), have brought enlarged consideration 
of the role of wage incentives in increasing productivity and in strengthen- 
ing the national economy. Furthermore, in a report by an Anglo-American 
Productivity Team sponsored by the Trades Union Council, note is made 
of the paradoxical situation that, while incentive systems are less common 
in Great Britain than in the United States, American trade unions are less 
concerned about the need for increased productivity than are trade unions 
in Great Britain (99). 


Observations made by the author in the U.S.S.R. in 1935 as well as 
recent reports indicate that wage-incentive payments are more prevalent 
in communistic than in other countries (105). In the early years of the 
Soviet Union, emphasis in the drive for increased labor productivity was 
placed upon propaganda appealing to the workers' enthusiasm to their 
"desire to build Socialism" leading to the creation of a corps of Sta- 
khanovites or "shockworkers" who form an integral part of the production 
system (52). Since approximately 1930, the tendency has been to lay in- 
creasing stress on differential-wage-scale and piece-work methods of wage 
payment ( 12) . Piece-work-wage plans and efficiency bonuses, without any 
guaranteed minimum wage, now constitute the basis for the compensa- 
tion of workers in industry, agriculture, and government (34). Wage- 
incentive plans have been extended to the point, where, for example, 
according to Kuznetsov, head of the trade unions, the piece work system, 
promoted by Stalin, accounted in 1949 for 92 per cent of wages in one 
industry and for similarly high percentages in other industries. As indi- 
cated by Newman, "this system of pay, once denounced by the Russians 
as the sweat shop and the worst form of capitalistic exploitation, has be- 
come the basic wage system of the Soviet Union" (65). 

In passing, it is interesting to note that the introduction of differential 
wage rates was severely criticized by members of the Communist Party 
who felt that this represented a return to the pernicious practice of capi- 
talistic societies a serious deviation from the fundamental principle of 
Communism voiced by Marx in his famous statement "from each ac- 
cording to his abilities, to each according to his needs." To such objec- 
tions the leaders of the Soviet Union have replied with striking facts con- 
cerning the adverse effect of equalitarianism in wages upon production 
and "labor fluctuation/' i.e., turnover (33); with labored reinterpretations 
of Marxism (93), and with restrictive and punitive measures against 
critics of Party policies. 

Along with such measures there has been much propaganda, directed 


at workers of the U.S.S.R. and other countries, emphasizing the sup- 
posed contrast between the harmful exploitation of labor through piece 
work in capitalistic countries and its beneficial effect upon the welfare of 
workers in the Soviet Union (90). Characteristically, the dialectical shift 
from the basic tenet of Marxism finds expression in the pronouncement, 
by Stalin, that "equalization in the sphere of demands and personal life 
is reactionary, petty bourgeois nonsense, worthy of a primitive ascetic 
sect and not of a socialistic society organized in a Marxian way" (94). 
Thus, Stalin rephrases Marx's statement to read "from each according to 
his capacity; to each according to his work/' 

Effectiveness of Wage Incentives 

Employers throughout the world continue to press for the more ex- 
tended use of wage-incentive plans. The drive for such plans is based pri- 
marily upon the belief that wage incentives lead to (1) greater output; 
(2) lower cost per unit of production; (3) higher earnings for employees 
and supplementary advantages in the way of reduced absenteeism, easier 
supervision, etc. (80, p. 10). According to Roy (84a), because of ad- 
ministrative costs and the loss of production resulting from the with- 
holding of effort by employees, wage incentive plans may actually lead 
to increased costs. However, reports from industry and surveys by gov- 
ernment bureaus generally attest to the effectiveness of wage-incentive 
plans in increasing productivity and achieving other objectives. A gov- 
ernment survey of 514 wage-incentive plans in the United States, re- 
ported in 1945, showed that, under such plans, production increases aver- 
aged 38.99 per cent; unit labor costs were decreased on the average by 
1 1.58 per cent, and average take-home pay was upped 17.56 per cent (66) . 

According to a survey of 48 companies manufacturing a wide variety 
of products and employing 14,000 workers in the New York area, the 
installation of wage-incentive plans resulted in production increases rang- 
ing from 3 to 103 per cent, averaging 27.9 per cent above "standard" and 
41.5 per cent above "past performance" ( 16) . Similar results are reported 
from a survey covering 62 companies in the Chicago area. Achievements 
relating to specific manufactured products are illustrated in Table 2. 

An example of the long-term effect of a wage-incentive system is found 
in the experience of the Shartle Brothers Machine Co., Division of the 
Black-Clawson Co. Here plant-wide wage incentives, both direct and 
indirect in character, have been applied for 11 years to all 350 shop em- 
ployees handling a great variety of parts and small quantity runs. The 
returns, in the form of greater production and of higher earnings for the 
workers, are shown in Figure 3. Records kept by the company show a 


TABLE 2. Results from Wage Incentives (from Nation's Business, 1945) 

Type of 

in pay, 
per cent 

in output, 
per cent 

Drop in unit 
labor cost, 
per cent 

Switch gears 



Gears and motor parts 




Fans and blowers 



Condensers, radio 







Wood products 




Corn products 




Radio and radar 




Pistol belts 



Canned foods 




Aluminum castings 




saving of 1,055,000 man-hours of labor in an 11-year period in this firm, 
where "workers asked to be time studied so that they, too, could earn 
bonuses" (37). 
The operation and effects of a newly established wage-incentive system 








Figure 3. Shartle Brothers Machine Co., Division of the Black-Clawson Co.; 
Eleven Years of Experience with Wage Incentives * 

* By permission, from R. K. Hovel, "Our eleven years with incentives/' Factory 
Mgmt. Mainten., (November) 1948, 106, 87-88. 



















in a plant is found in the experience of the Ecorse Division of the Murray 
Corporation of America, employing 1,000 workers in producing steel 
frames for motor vehicles (58). Here, a plan installed with the co- 
operation of the plant union (Local 2, UAW, CIO), led in a short time 
to production gains which averaged 16 per cent throughout the plants. 
According to the report on the experience of the Ecorse plant, "the aver- 
age would be even higher if certain old equipment could be replaced, and 
if the union had not imposed a ceiling of 25 per cent maximum above 
standard a restriction that reflects union qualms lest extraordinary indi- 
vidual earnings foment dissension" (58). Furthermore, contrary to ex- 
pectations, accident rates fell subsequent to the installation of wage in- 
centives, although, in contrast to the experience of other plants, there 
was no appreciable effect in the way of decreased absenteeism. Increased 
co-operativeness with supervisors was observed in the plant. From week 
to week, 94 per cent of direct workers showed an average increase in 
earnings of 19 per cent, while indirect labor showed an average gain of 
12 per cent in earnings. 


Reports such as those cited above suggest that wage incentives con- 
tribute to increased productivity. However, in most instances wage- 
incentive plans are introduced under conditions which make it difficult 
to evaluate the exact contribution of the wage incentive itself. Taylor is 
quoted as emphasizing the need of dealing with one variable at a time 
for "holding surrounding or accompanying conditions constant and uni- 
form while one variable is experimented with" (13, p. 247). In practice, 
the installation of a wage-incentive plan is generally accompanied by 
other changes in working conditions, personnel policies, and practices 
which are frequently major in character. For example, the introduction 
of wage incentives in the Ecorse plant was accompanied by the formula- 
tion of new "elemental standards/' probably necessitating changes in 
work methods and routines; by changes in shifts and in the length of the 
work week and, apparently, by a change in the over-all labor situation in- 
volving close co-operation between management and the labor union in 
the application of time-and-motion-study techniques. There are many 
instances in which the installation of a wage incentive has been ac- 
companied by changes in training programs and training methods; by 
improved procedures for evaluating employee performance, and even 
by higher selection standards and improved selection methods T (21, 

7 An interesting treatment of one phase of this problem is found in studies by 
Rothe (82, 83, 83a), who presents hypotheses bearing upon methods for isolating the 
specific effects of a wage plan in the "incentivation of workers." 


51). As a result, management and industrial engineers have frequently 
been unable to present clear-cut and unequivocal evidence as to the 
specific effects of wage incentives. 

The Psychologist Looks at Wage- Incentive Methods 

The evaluation of the motivating potential of wage-incentive plans is 
further complicated by the fact that there has been widespread failure to 
develop and adhere to scientific methods in the construction of wage- 
incentive systems (31, 34a, 46, 120). As suggested by Uhrbrock (101) in 
1935, the findings and techniques of psychology, physiology and other 
sciences pertinent to the consideration and evaluation of variables which 
enter into the construction of a wage-incentive plan have generally been 
neglected by exponents of scientific management, who formulate and use 
"laws" and "principles" without supplying data adequate for their sup- 
port or for their verification by means of independent investigation ( 101 ) . 
Particularly striking are flaws in the design and application of time-study 
techniques used in setting production standards. A review of this situation 
by Presgrave, in 1945, leads to the conclusion that "Practitioners have 
done virtually nothing to justify time study from scientific, sociological, 
physiological or psychological grounds, having been content to rest their 
case on the one proved fact that through their techniques they have been 
'preeminently successful in increasing output and decreasing costs' " (72, 
p. 1). Looking at wage-incentive methods from the viewpoint of the 
union leader, and on the basis of an exhaustive analysis of methods, Gom- 
berg also concludes that "thus far time study techniques can make no 
claim to scientific accuracy. They are at best empirical guides to setting 
up a range within -which collective bargaining over production rates can 
take place" (31, p. 170). 


The objective of time study is to arrive at a standard time which is used 
as a base in formulating production standards and in constructing the 
wage incentive plan (95a) . It is apparent that this calls for the completely 
dependable measurement of time intervals. Considerable dependence is 
still placed upon stop-watch readings in arriving at standard times. Fre- 
quently such standards are developed on the basis of a small number of 
observations made by a single time-study man. Questions can readily be 
raised concerning the reliability of the data obtained under such condi- 
tions (42a, 85) . Furthermore, methods used in selecting or computing the 
standard time for each element of a job are far from uniform. According 
to one procedure, the "time that recurs at least one third of the total 
number of readings, or if no such number is found, the number which 


with itself and any number less than itself occurs at least one-third of the 
total number of readings" (36) is taken as the standard time. In the work 
of other wage-incentive engineers, so-called "abnormal" times are elimi- 
nated largely on the basis of the opinion of the single observer (43) . Time- 
study data taken from different sets of observations by different observers 
are sometimes combined into synthetic studies. There is a serious question 
as to whether such data can be combined satisfactorily unless the "per- 
sonal equations" of different observers are known and compensated for. 

Added to such sources of variability are the estimates to thousandths 
of a minute made from stop watches with dials showing hundredths of 
a minute (28), and the distinct possibility that the stop watches used 
by different observers are far from uniform and accurate. The question 
has also been raised as to whether minute, elemental times no matter 
how accurate can justifiably be added to provide a true representation 
of the job as a "whole" (1, 29a). Furthermore, procedures used by 
some methods engineers involve no direct timing of the different ele- 
ments of the job, but depend in setting standards upon a so-called "law 
of motion," such as Segur's, which states, without adequate substantia- 
tion (87, p. 229 ff.), that "within practical limits the times required by 
all expert workers to perform true fundamental motions are constant" 
(la, p. 178). 

Such references to "what is wrong" with time-study practices are not 
intended to suggest that industrial engineers are unaware of these prob- 
lems (42a) . Thus, considerable attention has been directed to such ques- 
tions as whether "continuous" or "snap back" operation of the stop watch 
yields more reliable and more accurate measures (31, 72) . In the effort to 
overcome errors and improve reliability in timing, consideration has been 
given to the use of multiple observations. In considering this approach, 
Mundel recommends a statistical criterion for decisions as to the number 
of readings. According to the latter, "A reasonable limit to place on the 
number of readings would be to require a sufficient number such that 
the chances are 95 out of 100 that we are within 5 per cent of the true 
average for the element for the pace at which it was performed. Some may 
prefer a looser criterion of 95 chances out of 100 [that the readings are 
within] 10 per cent. In this latter case, the odds may be restated at 68 
out of 100 of being within 5 per cent" (61, p. 295). However, research 
by Leng, cited by Morrow (60), suggests that practical difficulties may 
exist in the satisfaction of such a criterion since it was shown that from 
14 to 75 observations, with an average of 28, were required to obtain a 
reliable reading to the nearest .01 minute. 

An analysis of time-study methods and findings leads Presgrave (72) 
to the assumption that the average element observed in time study is 


of .10 minute duration. An error of .01 minute thus represents an error 
of 10 per cent in the measurement of the average elemental time. Accord- 
ing to Gomberg, an error of this magnitude consumes practically the 
complete tolerance permitted by the collective-bargaining wage decre- 
ment or increment. "Obviously," he adds, "if after months of negotia- 
tions ... a settlement has been reached involving a 10 per cent change 
in the basic rates, neither management nor labor is prepared to sacrifice 
its respective rights to the blind operations of a technique of questionable 
accuracy" (31, p. 14). Such considerations have led to the view that time 
recorders should be substituted for stop-watch readings in making time 
studies. Studies of kymograph recordings have indicated that high levels 
of reliability can be achieved through the use of this device (7). A study 
by Leng, cited by Gpmberg, provides evidence of the superior accuracy 
of Marstochron as compared with stop-watch readings (31, p. 53ff.). 
Micromotion study, originally sponsored by the Gilbreths (26, 27, 29), 
involving the use of specially fitted cameras to produce chronocyclegraphs 
and stereochronocyclegraphs, results in increased precision of time 
measures, although other timing devices, such as the decimal camera and 
the Marstochron are preferred by other specialists in time study (96). 
On the other hand, Abruzzi reports that "there is not much difference 
between the stop-watch and the marstochron when the stop-watch is 
used by an experienced observer" (1, p. 244). The latter also refers to 
practical considerations which favor the continued use of the stop-watch. 
Such considerations have also been noted by Presgrave who, while ques- 
tioning stop-watch measures smaller than 1/1 00th of a minute, also con- 
tends that instruments of the type referred to above find their true use 
not in industry but in the laboratory (72, p. 50) . 


Ratings and Leveling. A common requirement in time study is that 
the standard time represent the performance of a man of average skill 
applying average effort under normal conditions (47, 95a). In arriving 
at such standards, it would seem necessary to establish objectively the 
fact that operators observed in time studies are actually men of average 
skill working with average effort. In practice, however, time studies do 
not involve operators selected with reference to distributions of workers 
with respect to such variables. Instead, operators are chosen on the basis 
of intelligence, cooperation, good will or even because they are superior 
operators and adjustments are made of observed elemental times in 
arriving at a standard time supposedly representative of the performance 
of an average man (68). This involves a process, commonly known as 
rating, defined by the National Committee on Effort Rating of the So- 


ciety for the Advancement of Management as "the process during which 
a time study engineer compares the performance of the operator under 
observation, with the observer's own concept of normal" (31, p. 131). 

There are wide differences among the methods used for adjusting the 
observed time to arrive at a standard appropriate to the average or "nor- 
mal" employee (45). Frequently employed is a leveling procedure, ad- 
vocated by Lowry, Maynard and Stegmerten, under which average time 
taken to complete task elements subject to control by the operator is 
multiplied by a leveling factor derived from ratings on skill, effort, con- 
sistency, and conditions (47). In still another application of ratings, 
recommended by Presgrave, only ratings on effort are used in the adjust- 
ment or "leveling" of time-study data (72), while Mundel takes the 
position that ratings on two factors, namely pace and job difficulty, be 
used in the adjustment of elapsed time (61a, 61b). 

Range of Individual Differences. It is evident, as has been made par- 
ticularly clear in critiques of time-study methods by Uhrbrock (101 ) and 
by Gomberg (31), that serious questions can be raised concerning rating 
procedures. Neglect of a scientific approach is illustrated in the varieties 
of assumptions made with respect to the range of individual differences in 
variables rated by time-study engineers. For example, a correspondence 
course for training methods engineers, cited by Uhrbrock, takes the posi- 
tion that the spread between "super skill" and "average skill" is 15 per 
cent (100). On the other hand, as a result of a study of 250 men who 
bored holes in small pieces of iron, Younger concludes that the "average 
man is 70 per cent slower than the skilled or fast man," and proceeds to 
subdivide the distribution of over-all times, on the basis of judgment, 
into zones of very fast, fast, average, slow, and very s/ow men (119). 
Weights for individual rating factors under still another plan range from 
0.22 to +0.15 in the case of skill, and from 0.04 to +0.04 in the case 
of consistency, and such weights (added to 1.00) combine to give a range 
of 0.50 to +1.38 for the leveling factor derived from the algebraic sum- 
mation of ratings on four such variables (47). 

Basic psychological data on the range of individual differences by Hull 
(38), Wechsler (111, 112) and others (82, 83), appear to have been 
largely overlooked in the application of rating factors. There is, it is true, 
growing recognition on the part of time-study engineers of the importance 
of such facts, and of the need for determining the limits of adjustment 
by means of appropriate research (47, 72). Barnes (4, p. 253), for ex- 
ample, illustrates the possibility of applying research techniques to such 
problems in an experiment involving 121 girls operating semiautomatic 
lathes. Here, under conditions which obviated the possibility that they 
were "pegging" production because of fear of a cut in the job rate or for 


other reasons, maximum output proved to be 104 pieces per hour and 
minimum output 51 pieces, a ratio of 1 to 2.04 which is in line with these 
reported in studies of aptitudes and skill by psychologists. Nevertheless, 
much still remains to be done both by psychologists and engineers in ac- 
cumulating data on output under varying conditions of skill, effort, and 
the like, as a basis for arriving at firm principles with respect to the ranges 
of variation to be taken into consideration in the application of ratings to 
particular work situations (87, p. 227 ff.). Research is also needed to 
examine further the possibility of eliminating or minimizing ratings 
through the application of a statistical approach, as suggested by Lorge 
(46) in 1935, and in a more recent analysis of time-study methods by 
Gomberg, who considers the possibility of applying "probability reason- 
ing" to the field of time study and of treating the concept of "average" 
or "normal" as "some quantitive characteristic of a parent population" 
(31, p. 133). 

The Reliability of Ratings. Of importance in the evaluation of time- 
study methods and findings is the distinct possibility, early recognized 
by Earth (8) and other pioneers in this field, that equally competent time- 
study engineers using the same rating factors might not agree in their 
evaluation of an operator's performance. Furthermore, there may be in- 
consistencies in ratings made by the same analyst from day to day. In- 
volved here is the recurring question of the reliability of subjective esti- 

In a study bearing upon this problem, reported by Uhrbrock (101), 
six judges, using a 10-point scale, independently rated 23 men on quality 
of output, sometimes used as a basis for evaluating skill. Correlations be- 
tween the ratings of two judges proved to be approximately .40, and of 
two independent sets of three judges, .63. Neither level of agreement can 
be considered as satisfactory, particularly in a situation where the size of 
the pay envelope and possibly also the tenability of a negotiated wage rate 
are affected by the ratings. 

Basically, the problems of reliability arising from the use of rating 
factors in time study are the same as these which arise wherever ratings are 
used. As a result, findings on the influence of the number of judges, on 
the effect of the number of variables rated, and the like, drawn from re- 
search in the field of merit rating (11, 70) and job evaluation (64, 106), 
have a bearing on the problems of reliability in rating or leveling. There 
is the possibility that reliability can be increased through the use of films 
as a substitute for direct observation of the operator (61b) . Nevertheless, 
in spite of some research in this area, the problem of augmenting the 
reliability of rating in time study needs to be solved as a necessary step 
in improving the construction and administration of wage-incentive plans. 


Recent reports indicate that studies directed toward this end are in- 
cluded in the research plans of the Rating of Time Studies Project sup- 
ported with funds derived from the sale of the film series "Study on Per- 
formance Rating" (89), described on page 36. 

The Validity of Leveled Times. Closely allied to the problem of re- 
liability is that of the validity of time standards arrived at through the 
application of rating and similar methods. One aspect of this problem is 
illustrated in a study by Cohen and Straus, involving 21 workers on a fold- 
ing operation (12a). Each operator was rated on skill and effort by three 
time-study engineers. Leveling factors derived from the engineers' ratings 
were then applied to the elapsed time of each operator as determined by 
a motion-picture time study involving the use of a 16-frames-per-second 
constant speed camera. In applying a rating or leveling procedure, the 
time-study engineer assumes that the base or leveled time (actual time 
multiplied by leveling factor) established by observing one operator will 
be exactly the same as that arrived at by observing any other operator, 
in spite of actual differences between the operators in skill, effort, and the 
like. In the study under consideration, leveled times were found to vary 
between a minimum of 59.04 and a maximum of 124.70, a range far 
beyond the limits of acceptability in setting what is presumed to be a 
valid index of the performance of an average worker of average skill work- 
ing with average effort under normal conditions. 8 

This situation in time study, and also its practical implications, are 
further illustrated in a study, reported by Gomberg (31), in which seven 
jobs in a garment-making plant were analyzed independently by a com- 
pany and by a union time-study engineer. Differences in actual cycle time 
recorded by the two engineers were all smaller than 2 per cent. However, 
disagreements between time-study engineers with respect to the magni- 
tudes of the rating factors 9 were such that the standard time unit assigned 
after leveling by a firm engineer was lower for each job than that assigned 
by a union engineer. There immediately arises a question as to which of 
the two standards can be considered as the more accurate index of the 
actual performance of an average man, working with average effort, under 
normal conditions, on each of the jobs. The possibility also exists that 
neither may be valid, in the sense that neither typifies the performance of 

8 Ryan chooses to discuss the findings of this experiment as illustrating the low 
reliability characterizing time standards "arrived at through studies of different workers, 
either by the same or different rate-setters" (87, p. 218), although recognizing that 
the problem considered might with equal propriety be viewed as one of validity. 

9 Differences in the magnitudes of the rating factors ranged from a low of 3.48 
per cent (blind stitch bottom) to a high of 16.70 per cent (set sleeves), with the 
rating factor assigned by the union engineer being in every instance higher than that 
allowed by the firm engineer. 


the conceptualized average or normal worker on the job or orders the 
job in correct sequence with reference to the other jobs. 

The literature on time-study methods reveals a growing awareness of 
the fact that conventional rating methods may yield times which are not 
in accord with the criterion of "standard time" as defined by the time- 
study engineer. Mundel has suggested the use of motion films as an aid 
both in formulating the criterion and in increasing the accuracy of rating 
(61b) . Under this proposal, a series of films of operators working at differ- 
ent paces on simple jobs is prepared. One of these is selected by manage- 
ment, or through joint negotiations by labor and management, as present- 
ing the desired standard pace, that is, "the rate of activity representing 
100 per cent standard pace." Films showing step-by-step deviations (pref- 
erably of the order of 6 per cent) from the accepted standard pace are 
then prepared and used as a basis for rating the pace on an observed job. 
Comparison of pace on the observed job with the "step" films projected 
into a shadow box placed near the job, or direct comparison of the "step" 
films with a film of the observed job, are among the methods recom- 
mended for use in establishing the pace rating. In an experiment by 
Lehrer, cited by Mundel, a group of experienced time-study men making 
comparisons with "step" films were able to place 64 per cent of their rat- 
ings within 5 per cent of the correct value, in contrast to 19 per cent 
when conventional methods of rating were used by the same men. 

The experiment referred to above is of particular interest in comparing 
the accuracy, and in this sense validity, of one rating procedure as com- 
pared with another. Many more such comparisons are needed to establish 
the relative validity of time standards arrived at by different rating meth- 
ods and for different tasks through the use of the same method. An effort 
has been made to overcome variations in standards due to differences in 
methods through the development of a series of films, showing 24 sep- 
arate industrial, clerical and laboratory operations, designed to present 
"a uniform concept of a fair day's work" (89, p. 4). Prepared under the 
auspices of the Society for the Advancement of Management, the films 
are based on approximately 150,000 ratings by more than 1,800 time-study 
engineers in over 200 companies throughout the nation. Data obtained 
through the application of the methods commonly used by these en- 
gineers were combined into a single standard viz. "the allowed time for 
the average of qualified incentive operators" (121, p. 53). According to 
the SAM report, the film provides "a concept of normal that can help 
solve disputes about specific time standards" (89, p. 4). However, Gom- 
berg and others have raised questions both as to the acceptability of the 
methods used and as to the possibility of arriving at an accurate concept of 


"normal working speed" through a study in which only management 
engineers and no union engineers participated (121). 

Criticisms of this kind, as well as the general consideration of time- 
study practices, point to the necessity for examining the criteria employed 
in setting the goals in a time-study program. Involved here, as suggested 
by Ryan (87, p. 221 ff.), are questions bearing upon: the acceptability of 
absolute reference points in rating, such as "average" worker (47); a pace 
equivalent to walking at 3 miles per hour (72); "average person . . . 
working at a pace of 100/130 of the maximum pace that can be main- 
tained, day after day, without harmful physical effects" (61b), etc. Also 
involved are broader questions bearing upon what constitutes a "fair day's 
pay" (121), upon the role and objectives of incentives in a social order, 
and others which challenge the ingenuity of the social scientist, as well 
as that of the wage-incentive engineer. 


In addition to the application of ratings in setting time standards, fur- 
ther adjustments are generally made in the form of allowances to provide 
for personal needs, for delays over which the worker has no control, to 
prevent undue fatigue, and the like. Thus, in an illustration cited by 
Barnes (4), an average elapsed time for a work operation of 2.20 minutes 
is increased to a base or leveled time of 2.93 minutes by the application 
of skill and effort ratings, and then further increased to 3.10 minutes by 
the application of a 5-per-cent allowance for fatigue and personal needs. 

As in the case of rating factors, there are variations among time-study 
engineers with respect to the types of allowance and to the time allowed. 
Thus, in a study reported in 1928, allowances for fatigue were found to 
range from no adjustment at all to an allowance as high as 65 per cent 
(95 ) . A survey of time-study literature by Mundel, in 1950, shows a range 
of "personal allowances" from a maximum of 4 per cent under one system 
to a maximum of 50 per cent under another (61c). 

Studies of time taken by workers to attend to personal needs can furnish 
objective data for estimating the allowance for such needs. Similarly, data 
on delays caused by machine breakdowns and other factors not under the 
control of the worker can provide a factual basis for the delay allowance 
(1, p. 252 ff.) . Data from studies of work decrement during various parts 
of the day particularly at the end of the work shift have been used as 
a basis in deciding on the size of the fatigue allowance (4, 87). However, 
many problems, involving both the theory and practice of time study, are 
involved in setting allowances (41a). From the viewpoint of a scientific 
approach, arriving at allowances for fatigue constitutes a problem of par- 


ticular complexity, since studies by psychologists and physiologists have 
produced little in the way of usable techniques (other than output itself) 
for measuring fatigue. Furthermore, much difference of opinion still 
exists with respect to the very nature of fatigue and, in particular, as 
to whether reduced output (work decrement) provides an adequate 
measure of fatigue effects (8a, 31, 32, 63, 87). It has been suggested that 
this situation might be relieved by eliminating allowances for fatigue 
through the substitution of properly spaced rest pauses of appropriate 
duration. However, although attention was called to the problem by 
Uhrbrock in 1935, research is still needed to establish scientifically 
"whether or not it is most effective and desirable to add a short time al- 
lowance for fatigue to the time required to perform each element of work 
or to provide carefully spaced rest periods of several minutes' duration 
throughout the working day" (101, p. 21). 


The use of ratings or leveling factors and allowances of various kinds 
means that the standards used as a basis for wage incentives are derived 
from a series of measurements supplemented by qualitative observations 
and judgment. This situation was accurately described as early as 1921 
by an outstanding exponent of scientific management: "The time-study 
job," according to Spaeth, "is tosfet a fair wage on the basis of his observa- 
tions not only of time intervals, but with the addition of allowances for 
fatigue, delays, etc. The final result, therefore, whether with a stop watch 
or with a moving picture technique, is made up of a combination of some 
very accurate measurements and some guesses. Consequently, your best 
guess necessarily determines the order of accuracy of the final result" (92 ). 

There is evidence of growing recognition of this situation by the ex- 
ponents of time study. There are indications that the problems of reliabil- 
ity and validity are receiving increasing attention by time-study engineers. 
Developments in the field of time study, involving in some instances an 
effort to substitute statistically developed procedures for conventional 
time-study procedures (96), may reflect wider acceptance of an approach, 
favored by Lorge (46) and by Gomberg (31), which avoids or at least 
minimizes the use of subjective estimates in time study. However, in prac- 
tice today, time standards continue generally to represent figures derived 
from a combination of measurements made under more or less uniform 
conditions with allowances reflecting the judgment of time-study engi- 
neers. In other words, standards are generally not based solely on objec- 
tive, accurate, reliable, and verifiable data. This situation, according to 
Uhrbrock, has contributed to employee opposition to wage-incentive 
plans, since workers can justifiably raise questions concerning the judg- 


merits which are made, and become skeptical as to the exactness of the 
standards arrived at through the procedures ordinarily used in setting 
standards (101). Increasing demands by unions for time-study data pre- 
pared by the company, and the support of such demands by NLRB de- 
cisions, provide another practical reason for the improvement of methods 
to furnish data which will have a sound scientific basis and be uniformly 
meaningful to both the company and union representatives. 


It is evident, from the foregoing discussion, that much is needed in the 
way of technical improvements in methods used in the development of 
wage-payment systems to insure the optimal utilization of the wage in- 
centive in industry. However, in the opinion of many observers, the tech- 
nical improvement of wage-incentive systems represents only a partial 
answer to the effective motivation of employees. For example, according 
to Whyte, "Systems of financial incentive in industry yield a net gain in 
productivity, but most of them fail to release more than a small portion 
of the energy and intelligence workers have to give to their jobs/' This, 
he adds, is because "the setting of incentives has been traditionally con- 
sidered a technical engineering problem/' and there has been failure 
to recognize that "the problem has important human relations aspects 
that cannot be solved through the enforcement of any technical stand- 
ards . . ."(115, p. 79). 

The need for considering wage incentive systems as a problem in hu- 
man relations (as well as a technical problem) appears clearly in studies 
of employee reactions to wage incentives, discussed in Chapter 3. Such 
studies, and other research in industry, have also turned attention to the 
consideration of other factors, discussed in succeeding chapters, which 
operate alongside of, or apart from, the hope of financial reward in de- 
termining the levels of employee production, job satisfaction, and morale 
in the individual plant and in industry at large. 


1. A. Abruzzi, Work Measurement. Columbia University Press, 1952. 

la. L. Alford, Laws of Management Applied to Manufacturing. The Ronald Press 
Co., 1928. 

2. C. C. Balderston, Group Incentives. University of Pennsylvania Press, 1930. 

3. C. C. Balderston, "Possibilities and Limitations of Various Types of Profit Shar- 
ing/' Employee Benefit Plan Review (Spring Issue), 1948. 

4. R. M. Barnes, Motion and Time Study. John Wiley & Sons, Inc., 1949 (3rd 


5. R. M. Barnes, Work Measurement Manual. Wm. C. Brown Co., 1947. 

6. R. M. Barnes and N. A. Engbert, Bibliography of Industrial Engineering and 
Management Literature. Wm. C. Brown Co., 1946. 

7. R. M. Barnes and J. S. Perkins, "A study of the development of skill during 
performance of a factory operation/' A.S.M.E. Transactions, 1941, 63, 319-28. 

8. C. G. Earth, "Stop watch time studies: A symposium, II. A defense of the stop 
watch," Bull. Taylor Society, 1921, 6, 108-111. 

8a. S. H. Bartley and E. H. Chute, Fatigue and Impairment in Man, McGraw- 
Hill Book Co., 1947. 

9. C. E. Bedaux, "The Bedaux-Unit principle of industrial management/' /. Appl. 
Psychol 1921, 5, 119-26. 

10. L. E. Bethel, F. S. Atwater, G. H. E. Smith and H. A. Stackman, Industrial 
Organization and Management. McGraw-Hill Book Co., 1950, 609. 

11. H. Burtt, Principles of Employment Psychology. Harper & Brothers, 1942 
(Especially Chapter 12). 

12. W. H. Chamberlin, Russia's Iron Age. Duckworth, 1935, 61 ff. 

12a. L. Cohen and L. Strauss, "Time study and the fundamental nature of skill/' 
/. Consult. Psycho/., 1946, 10, 146-53. 

13. F. B. Copley, Frederick W. Taylor. Harper & Brothers, 1933. 

14. C. B. Corby, "Everyone gets a share of the profits/' Factory Mgmt. Mainten., 
1937, 95, 82-83. 

14a. H. J. Daigneault, "Sharing the profits of increased productivity" in Spot- 
lighting the Labor-Management Scene (Personnel Series No. 147). New York: 
American Management Assn., 1952, 9-16. 

15. R. W. Davenport, "Enterprise for everyman," Fortune, (January) 1950, 41, 55 ff. 

16. D. Despain, "Let workers writtf the pay checks/' Nation's Business, 1945, 33 

17. Z. C. Dickinson, Compensating Industrial Effort. The Ronald Press Co., 1938. 

18. A Digest of Selected Articles on Wage Incentive Plans (Industrial Relations 
Memo, No. 72). New York: Industrial Relations Counselors, Inc., 1944. 

19. V. K. S. Dixon, "The worker as a fiscal partner in industry/' Partners, 1951, 
5 (July-August), 10-14. 

20. M. Dobb, Wages. Cambridge University Press, 1941. 

21. C. R. Dooley, "What are the trends in wage incentive and other financial in- 
centives?," Personnel, 1935, 11, 120-22. 

22. H. Emerson, Efficiency as a Basis for Operation and Wages. New York: The 
Emerson Engineers, 1912 (3rd Edition). 

23. Experience of 123 Companies with Wage Incentive Plans (Special Investigation 
Report no. 523). Chicago: Dartnell Corporation. 

24. P. S. Florence, Labour. Hutchinson's University Library, 1950. 

25. H. L. Gantt, Work, Wages, and Profits. New York: The Engineering Maga- 
zine Co., 1913 (2nd Edition). 

26. F. B. Gilbreth, Motion Study. D. Van Nostrand Co., Inc., 1911. 

27. F. B. Gilbreth and L. M. Gilbreth, Applied Motion Study. Sturgis and Walton 
Co., 1917. 

28. F. B. Gilbreth and L. M. Gilbreth, "Stop watch time study: A symposium; VI. 
Standard times for standard 'work elements/ " Bull. Taylor Society, 1921, 6, 

29. L. M. Gilbreth, "Time and motion study," Occupations, 1934, 12, 35-39. 


29a. J. J. Gillespic, Dynamic Motion and Time Study. Brooklyn: Chemical Publish- 
ing Co., 1951. 

30. C. S. Golden and H. J. Ruttenberg, The Dynamics of Industrial Democracy. 
New York: Harper and Brothers, 1942. 

31. W. Gomberg, A Trade Union Analysis of Time Study. Chicago: Science Re- 
search Assoc., 1948. 

32. W. Gomberg, "Measuring the fatigue factor/' Industr. Lab. Relat. Rev., 1947, 
1, 80-93. 

33. J. Grabe, New Methods of Work New Methods of Leadership. Moscow: 
Co-op. Pub. Soc. of Foreign Workers in U.S.S.R., 1933, 24-29. 

34. V. Gsovski, Elements of Soviet Labor Law (Bureau of Labor Statistics Bulletin 
No. 1026). U.S. Govt. Printing Office, 1951. 

34a. H. W. Haggard, A Physiologist Looks at Time Study, Motion Study, and 
Wage Incentive Methods (Management Series No. 18). New York: American 
Management Assn., 1937, 4-17. 

35. F. A. Halsey, "The premium plan of paying for labor," Economic Studies, 1896, 

36. C. C. Herrman, "The standard day's work," Industr. Mgmt., 1920, 59, 288-89. 

37. R. K. Hovel, "Our eleven years with incentives," Factory Mgmt. Mainten. f 
1948,106, 11,87-88. 

38. C. L. Hull, Aptitude Testing. World Book Co., 1928. 

39. "Incentive payments in Australian industry/' Bull. Industr. Psychol. Pers. Prac- 
tice, 1950, 6, 13. 

40. Incentives for Management and Workers. New York: (Production Series No. 
161) American Management Assn., 1945. 

41. "Incentives for pieceworkers/' Industr. Re/., 1931, 2, 446-47. 

41a. C. L. M. Kerkhoven, "Time study research," Advanc. Mgmt., 1952, 15, 11, 

42. C. Kress, "Profit sharing merits No. 1 spot on agenda of American Industry/' 
Employee Benefit Plan Review, (Mid-Year Issue) 1947. 

42a. I. P. Lazarus, "The nature of stop watch time study errors," Advanc. Mem*., 
1950, 15, 5, 15-16. 

43. W. O. Lichtner, "Time and job analysis in management V. How to work up 
and use time studies," Industr. Mgmt., 1920, 60, 130-39. 

44. J. F. Lincoln, Incentive Management. Cleveland: Lincoln Electric Co., 1951 

45. S. B. Littauer and A. Abruzzi, "A survey of time-study practices," Personnel 
1949, 26, 123-32. 

46. I. Lorge, A Statistician Looks at Time Study, Motion Study, and Wage- 
Incentive Methods (Management Scries No. 18). New York: American Man- 
agement Assn., 1937, 18-26. 

47. S. M. Lowry, H. B. Maynard and C. J. Stegemerten, Time and Motion Study. 
McGraw-Hill Book Co., 1940 (3rd Edition). 

48. R. P. Lynton, Incentives and Management in British Industry. Routledge and 
Kegan Paul, 1949. 

49. C. W. Lytle, Wage Incentive Methods. The Ronald Press Co., 1945. 

50. C. A. Mace, "Advances in the theory and practice of incentives," Occufi. Psy- 
chol., 1950, 24, 239-44. 

51. C. Madge, "Payment and incentives/' Occup. Psychol, 1948, 22, 39-45. 

52. S. K. Margold, "Economic life in Russia's orbit/' Harvard Bus. Rev., 1950, 28, 


53. R. Marriott, "Socio-psychological factors in productivity,'* Occup. PsychoL, 
1951, 25, 15-24. 

54. H. B. Maynard, G. T. Stegemerten and L. Schwab, Methods Time Measure- 
ment, McGraw-Hill Book Co., 1948. 

55. Ministry of Labour Gazette, (October) 1947, 324-25; 332-33. 

56. Ministry of Labour and National Service, Industrial Relations Handbook. H. M. 
Stationery Office, 1944. 

57. Ministry of Labour and National Service, Wage Incentive Schemes. H. M. Sta- 
tionery Office, 1951. 

58. "More pay more production," Fortune, (September) 1943, 138 ff. 

59. A. H. Mogensen, Morion and Time Study. McGraw-Hill Book Co., 1932. 

60. R. L. Morrow, Time Study and Motion Economy. The Ronald Press Co., 1946. 

61. M. E. Mundel, Motion and Time Study. Prentice-Hall, Inc., 1950. 

61a. M. E. Mundel, "Time study research, Part I Rating/' Advanc. Mgmt, 1950, 

15,6,13-16. , 
61b. M. E. Mundel, Time study research, Part II Rating/' Advanc. Mgmt., 

1950, 15,8,20-22. 
61c. M. E. Mundel, "Time study research, Part IV Allowances," Advanc. Mgmt., 

1950, 15, 10, 22-24. 

62. M. E. Mundel and A. J. Keim, "Proceedings of second annual Purdue motion 
and time study work session," Misc. Bui. of Div. of Tech. Ext., Purdue Uni- 
versity, (April) 1945. 

63. National Research Council Committee on Work in Industry, Fatigue of Workers. 
Reinhold Publishing Corp., 1941. "^ 

64. J. Otis and R. H. Leukart, Job Evaluation. Prentice-Hall, Inc., 1948 (Especially 
Chapter 10). 

65. J. Newman, "Soviet worker a serf in red company union/' Phila. Inquirer, Nov. 
4, 1949. 

66. "Pay plans for higher production/' Modern Industry, (April 15) 1946, 51 ff. 

67. Personnel Activities in American Business (Studies in Personnel Policy No. 86). 
New York: National Industrial Conference Board, 1946. 

68. J. A. Piacitelli, "Selection of workers for observation," Bull. Taylor Society, 1921, 
6, 113-14. 

69. 'The pitfalls of profit sharing," Fortune, (August) 1951, 104 ff. 

70. A. T. Poffenberger, Principles of Applied Psychology. Appleton-Century. 1942 
(Chapter 14). 

71. Practical Uses of Wage Incentives (Production Series No. 150). New York: 
American Management Assn., 1944. 

72. R. Presgrave, The Dynamics of Time Study. McGraw-Hill Book Co., 1945. 

73. Productivity and the Worker. Princeton: Opinion Research Corporation, 1947. 

74. Profit-Sharing A Study of Results of Overseas Experience. Melbourne: Depart- 
ment of Labor and National Service, 1947. 

75. Profit Sharing for Workers (Studies in Personnel Policy No. 97). New York: 
National Industrial Conference Board, 1948. 

76. Pro/it Sharing for Executives (Studies in Personnel Policy No. 90). New York: 
National Industrial Conference Board, 1948. 

77. H. Reitinger, Sound Wage Incentives. New York: The Emerson Engineers, 1941. 

78. Report of Building Industry Productivity Team. H. M. Stationery Office, 1950. 

79. Report of Petrol and Oil Productivity Team. H. M. Stationery Office, 1950. 


80. A Review of Wage-Incentive Practice. Melbourne, Australia: Department of 
Labour and National Service, 1949. 

81. L. S. Ronin, "Planned economy in operation in the Soviet Union/' in World 
Social Economic Planning. The Hague: International Industrial Relations Insti- 
tute, 1932, 341-77. 

82. H. F. Rothe, "Output rates among chocolate dippers/' /. Appl. Psychol., 1951, 
35, 9497 

83. H. F. Rothe, "Output rates among machine operators/' /. Appl. PsychoL, 1947, 
31, 48489. 

83a. H. F. Rothe, "Output rates among butter operators/' J. Appl. PsychoL, 1946, 

84. R. L. Rowe, "Profit sharing in industry/' Harvard Bus., Rev., 1949, 27, 559-84. 
84a. R. H. Roy, "Do wage incentives reduce costs?" Industr. Lab. Rel. Rev., 1952 5 


85. E. B. Royer, "How many observations are necessary in setting wage incentive 
standards?/' Personnel, 1937, 13, 137-39. 

86. F. Russon, Bonusing for Building and Allied Trades. Birmingham: Norman 
Tiptaft, 1950. 

87. T. A. Ryan, Work and Effort, The Ronald Press Co., 1947. 

88. J. Scanlon, Adamson and His Profit-Sharing Plan (Production Series No. 172). 
New York: American Management Assn., 1947. 

89. H. F. Smiddy, "Society launches major research activity," Advanc. Mgmt., 1952, 
17, 7, 35. 

90. D. A. Schmidt, "Czechoslovakia: Lesson to the West," New York Times Maga- 
zine, July 31, 1949, 7ff. 6 

91. Some Problems in Wage Incentive Administration (Studies in Personnel Policy 
No. 19) New York: National Industrial Conference Board, 1940. 

92. R. A. Spaeth, "Stop watch time study. A symposium, VII. Time study and 
human standardization," Bull. Taylor Society, 1921, 6, 114-117. 

93. J. Stalin, On Technology. Moscow: Co-op. Pub. Soc. of Foreign Workers in 
U.S.S.R., 1932, 55-58. * 

94. J. Stalin, Problems of Leninism, 1938, 583. Cited from V. Gsovski, Elements 
of Soviet Labor Law (Bureau of Labor Statistics, Bulletin No. 1026) U.S. Govt 
Printing Office, 1951, 2. 

95. N. E. Stearns, "Computation of allowances," BuU. Taylor Society, 1928, 13, 

95a. C. J. Stegemerten, "Control of production through time standards," in 
H. B. Maynard (Ed.), Effective Foremanship. McGraw-Hill Book Co., 1941, 

96. L. A. Sylvester, The Handbook of Advanced Time-Motion Study. Funk & 
Wagnalls Co., 1950. 

97. W. F. Taylor, The Principles of Scientific Management. Harper and Brothers, 
1911, 129-30. 

98. K. H. Thompson, Profit Sharing. Harper and Brothers, 1949. 

99. Trade Unions and Productivity. British Trades Union Council, 1950. 

100. Training Course, A. Motion and Time Study. Pittsburgh: Methods Engineering 
Council, 1934, Section VI, 3. * * 

101. R. S. Uhrbrock, A Psychologist Looks at Wage Incentive Methods (Institute of 
Management Series, No. 15). New York: American Management Assn., 1935. 

102. U.S. Monthly Labor Review, November, 1947, 535-38. 


103. V. V. Veysey, "Wage incentives and war production," Pen. /., 1944, 23 (Jan- 
uary), 42-45. 

104. M. S. Viteles, Industrial Psychology. W. W. Norton & Co., Inc., 1932. 

105. M. S. Viteles, "Industrial psychology in U.S.S.R.," Occup. Psychol, 1938, 12, 

106. M. S. Viteles, "A psychologist looks at job evaluation," Personnel, 1941, 17, 

107. Wage Incentives. (Information Bulletin No. 5). New York: National Associa- 
tion of Manufacturers, 1941. 

108. "The wage incentive," Fortune, (September) 1943, 136ff. 

109. Wage Incentive Practices. (Studies in Personnel Policy No. 68). New York: 
National Industrial Conference Board, 1945. 

110. Wage Payment Systems. (Studies in Personnel Policy No. 91). New York: Na- 
tional Industrial Conference Board, 1948. 

111. D. Wechsler, "The range of human capacities," Sci. Monthly, 1930, 31, 35-39. 
112 D Wechsler, The Range of Human Capacities. The Williams and Wilkins 

Co, 1952. 

113. A. Ramond, "Philosophy for practice of wage incentives," Proceedings, Third 
Annual Industrial Relations Seminar. Washington, D.C.: Radio Manufacturers 
Assn., 1947, 38-50. 

114. W. H. Wheeler, "Profit-sharing the spark that jumps the gap," 19th. Pacific 
Coast Management Conference of the California Personnel Management As- 
sociation, 1948. 

115 W F Whyte, "Economic incentives and human relations," Harvard Bus. Rev., 
1952, 30, 73-80. 

116. W. F. Whyte, "Incentives fof productivity; the Bundy Tubing Company Case," 
Appl. Anthrop. f 1948, 71, 1-16. 

117. W. F. Whyte, "Union-management cooperation: A Toronto case," Appl. 
Anthrop. t 1947, 6, 1 ff. 

118 D. Yoder, Personnel Management and Industrial Relations. Prentice-Hall, Inc., 
1947, 380 ff. 

119. J. Younger, "Finding the average man," Amer. Machinist, 1931, 75, 879. 

120. S. G. Yulke, "Incentive and work standard grievances" in Production Policies 
for Increased Output (Production Series No. 19). New York: American Man- 
agement Assn., 1947, 27-36. 

121. "The fair day's work can time-study experts measure it objectively?" Fortune, 
(May) 1951, 52-54. 

The Worker and 
Wa$e Incentives 

"I HEARD the company hired one of those efficiency experts. 
Up to now this has been a pretty good place to worfe, but he'll probably 
ruin everything!" (12, p. 87). This thought, expressed by workers of one 
company when they learned that an industrial engineer had been en- 
gaged to install an incentive system, typifies the suspicion and distrust 
with which wage-incentive plans are frequently viewed by workers. 

In some instances, suspicion has developed into direct hostility to wage- 
incentive systems and to the way in which they have been administered in 
industry. Such antagonism has been reinforced by a systematic policy of 
opposition on the part of some unions, and by skepticism with respect to 
the merit of wage-incentive plans on the part of other elements of the 
population (23, 24). 


Opposition to wage-incentive systems appeared in this country almost 
immediately after their introduction by Taylor. According to his biogra- 
pher, Taylor continually fought against the opposition of workers when 
he introduced his methods at the Midvale Steel Co. and, as a young fore- 
man, Taylor was "frequently threatened with violence, and the fight in- 
creased in bitterness; the men resorting to such things as the deliberate 
breaking of machines, and their foremen to large-scale fining and firing" 
(4, p. 5). 

Antagonism to the principles and practices of scientific management 
was emphasized in a report of a Congressional committee, in 1912, which 
described "stimulation" as the "crux of the difficulty in the introduction 
of the Taylor or any other system of so-called scientific management"; de- 



clared that no such system should be introduced without prior consulta- 
tion with workers or imposed upon an unwilling working force, and recom- 
mended that stop-watch time-studies should not be made without the 
consent of the workmen involved (26). Sufficient public support was 
aroused to lead to the introduction, in 1914, of an amendment to a bill 
regulating the method of directing the work of government employees 
which would have prohibited time studies of Federal employees (27). 
While the bill was not actually enacted into law, the fact that it was 
introduced illustrates the hostility to wage-incentive methods and plans 
which developed simultaneously with the use of scientific management. 


Interviews with workers have shown that, even in the absence of overt 
hostility, wage-incentive systems frequently receive merely passive ac- 
ceptance as a necessary or inescapable evil (1, 5). It has also become ap- 
parent that failure to understand the plan a lack of clear perception 
of the goals is, at times, at the root of such attitudes. Difficulties of this 
sort, resulting from defects in the engineering and administration of wage 
incentive plans discussed in Chapter 2, are illustrated in a worker's com- 
ment, cited by Uhrbrock, to the effect that "Well, the main trouble with 
that Wage-Incentive as I see it is that it's too complicated for the ordinary 
working man to understand" (25, p. 9) . Another illustration comes from 
a study conducted in two large British engineering (metal-trade) factories 
on behalf of the Industrial Health Research Board. "We've been to ask 
them to explain" said one of the female operatives, "but the gentlemen 
in the office say it's in decimals which we couldn't understand, and that 
we must go back to work and not make a fuss" ( 5, p. 58) . Workers some- 
times fail to heed such advice and, instead, develop deep-seated antago- 
nisms to wage incentives. The crucial character of this situation is well 
illustrated in the conclusion, reached by union leaders, that, "if the in- 
dustrial engineers of today would tear the veil of secrecy and mysticism 
from the application and operation of their innumerable wage incentive 
plans, we might be able to achieve a mutually satisfactory acceptance of 
wage incentives . . . Any wage incentive plan, any standard that has 
been set, that will not stand frank open discussion and complete explana- 
tion to the people concerned, leaves a lurking suspicion of unfairness and 
dishonesty" (8, p. 56). 

Employee Attitudes toward Incentive- and 
Non incentive- Wage Systems 

Information concerning workers' appraisals of wage incentives is avail- 
able from surveys which have dealt with employee attitudes toward in- 


centive as compared with nonincentive systems of wage payment. Illus- 
trative of studies bearing upon this question is a survey conducted in 
1945, by the Opinion Research Corporation (29), involving a nation-wide 
sample of manual workers in manufacturing plants located in cities with 
a population of 25,000 and over. All of the workers involved in the survey, 
totaling 919, were asked, "On a job which could be paid by either piece 
rate or hourly rate, which one would you rather work on?" The results, 
summarized in Table 3, show that ( 1 ) 24 per cent of the workers on an 
hourly rate say they would prefer to be paid by piece rate; (2) 22 per cent 
of workers on piece rate and 39 per cent working under other incentive 
plans would prefer to be paid at an hourly rate; (3) workers tend to prefer 
the wage plan under which they are actually working; (4) preference for 
hourly rate is higher among union members than among nonunion mem- 
bers of the working force. 

Employees working under a piece-rate plan were asked, "If you had an- 
other job would you like to be under the piece-rate system or some other 
system?" Those working under some other incentive-pay system were 
asked, "If you had another job, would you like it to be under a system 
like the one you're now on, or some other system?" The members of both 
of these groups were asked, "What are some of the things you like about 
your present system?" The replies, summarized in Table 4 (top section), 
show again that most workers express preference for the system under 
which they are already working. The table (bottom section) also shows 
what workers like about a system of wage payment in which earnings are 
related to production. 

TABLE 3. Replies to Question: "On a job which could be paid by either piece 

rate or hourly rate, which would you rather work on?" (after Public 

Opinion Index for Industry, Opinion Research Corporation, 1946) 

No. of Per cent who prefer 

Mfg. man- Piece Hourly Don't 
ual workers rate rate know 

Total 919 36 61 3 
Paid by: 

Hourly rate 658 24 73 3 

Incentive plan 131 57 39 4 

Piece rate 130 75 22 3 

Union status: 

No union where work 220 43 53 4 

Have union 699 34 63 3 

Members 597 33 65 2 

Nonmembers 102 35 54 11 



TABLE 4. Employees 9 Attitudes toward Wage Incentives (after Public 
Opinion Index for Industry, Opinion Research Corporation, 1946) 

Paid by 

Incentive Piece 


Prefer new job on present system 
Another system 
Don't know 

131 130 

Per cent 
69 61 

30 35 

1 4 

What liked about present system 
Number preferring present system 

It's up to me what I make: Tests my ability; indi- 
vidual energy is rewarded; pav doesn't depend on 
others; I get paid for what I ao 
I'm my own boss: No one hanging over me to see 
I work; I work fast or slow as I please 
Make more money than on hourly rate: If rate is 
set fairly you make a good income 
Sure of at least minimum pay: Not under strain 
because you know you'll get a certain amount 
Can figure pay easily: If yoaVe done so much 
work you know you'll earn so much; no argu- 
ment about amount coming 
Can increase earning power: If you stay on one 
job, get more skill, you can earn more 
Everybody works together: All pull together; if 
you get stuck someone helps you out; you have 
a feeling of co-operation, crew work 
Miscellaneous reasons 
No reason given 

91 79 

Per cent * 











* Percentages add up to more than 100 because some respondents gave 
more than one reply. 

According to Factory's fourth annual survey of What the Factory 
Worker Really Thinks, reported in 1947, also involving a nation-wide 
sampling of nonsupervisory factory employees, 30 per cent of those inter- 
viewed said they were paid on an incentive or bonus basis. Of these, 85 per 
cent of nonunion workers and 75 per cent of union workers, constituting 
79 per cent of all receiving incentive wages, stated that the incentive sys- 
tems in their plants were fair. Moreover, 51 per cent of workers not paid 
on an incentive basis are reported as saying that "they would like to work 
under such a system if it were fairly run/' According to the interpretation 
offered by the editors of Factory, the "workers who are not on incentive 


but want to be, constitute a real challenge to management and to those 
unions which stand in the way of introducing incentive systems" (33, p. 

Evidence that the wage incentive represents, at best, only a partial solu- 
tion of the productivity problem is found in a survey by the Opinion Re- 
search Corporation, conducted in 1949, involving 1,021 manual workers 
comprising a national sample of employees of manufacturing companies. 
As appears in Figure 4, 65 per cent of the workers agree that incentive 











Figure 4. Workers Acknowledge that Pay Geared to Output Raises Produc- 
tion but a Wide Majority Prefer to Work on Hourly Pay * 

pay results in the most output per man, but the same percentage prefer 
hourly pay to incentive pay. The ratio of preferences for hourly pay was 
found to be 74 to 20 in the case of hourly workers; 59 to 36 in the case of 

* By permission, from Public Opinion Index for Industry, Opinion Research Corpo- 
ration, 1949. 


workers already on incentive pay. These findings are interpreted as show- 
ing that "as useful as incentive pay is in raising output, it does not in itself 
solve the problem of obtaining workers 9 cooperation. In some circum- 
stances it may intensify that problem" x (19, p. 18). 


It is apparent that the findings and conclusions from surveys bearing 
upon employee attitudes towards wage incentives are not in complete 
agreement. Differences in population samples, in the forms of questions, 
and in other conditions discussed in Chapter 12, help to explain such 
discrepancies. The situation is further complicated by the possible effects 
of lack of direct knowledge on the part of the worker of wage systems 
other than that under which he is employed. In general, too, the worker's 
preference for the system under which he is working, whether it be an 
hourly rate or a wage incentive plan, may reflect the influence of a deep- 
seated human characteristic, viz., resistance to change the tendency (of 
each of us) "to fight against recognizing the necessity for changes that 
is apparent to everyone but ourselves" ( 14, p. 53) . However, both day-to- 
day observations in industry and systematic studies of output restriction 
indicate quite clearly that the sources of employee discontent with wage- 
incentive systems and their administration are both numerous and com- 
plex in character. 

Production is adjusted by the worker not only because of the human 
tendency to resist change, but also because of fear of rate cuts and resent- 
ment against speed-up; as a means of postponing layoffs; steadying em- 
ployment, and stabilizing earnings (15). In addition, such factors as 
grievances, "hard feelings" toward management, and discouragement un- 
questionably interfere with maximum production (3, 17). Opposition to 
increased productivity, in the face of wage incentives, may result from the 
belief that profits are too high (29, 33); from the popularly phrased argu- 
ment "Why should I break my neck so that the boss's wife can have a 
second fur coat?" (6, p. 82), reflecting the worker's opinion that if he 
produces more the company pockets the gain and he does not share fairly 
in the benefits of increased output ( 19) . An effort to protect the interest 
of less competent fellow workers is at the base not only of union policies 
with respect to the restriction of output, but of voluntary restriction on 
the part of the individual workers ( 1 ) . Lack of confidence in standards 
established by company rate-setters without direct participation of work- 
ers' representatives (29) is increasingly stressed, particularly by union 
leaders and members, as a reason for withholding effort under a wage- 
incentive plan (8). 

The fact that such feelings and sentiments determine the attitudes 

1 Not italicized in the original. 



of employees towards wage incentives is evidenced in the second annual 
nation-wide survey of what the worker really thinks conducted for Factory 
by the Opinion Research Corporation, in 1945 (32). As stated on page 
5, 49 per cent of all the manual workers questioned were of the opinion 
that "when a man takes a job in a factory he should turn out as much as 
he can' 9 The opinion that he should "turn out the average amount" was 
expressed by 40 per cent of all manual workers. Workers who gave the 
latter response were asked, "What do you think would happen if he 
turned out MORE than the average?" The replies, summarized in Figure 
5, reveal the extent to which some of the sentiments referred to above 
influence workers' reactions to wage incentives. 

Question: (addressed only to those who said "average amount"): What 
do you think would happen if he turned out more than the 


Management Would Raise 
Production Quotas 

It Would Be Unpopular 
with Other Workers 

Piece Rates Would 
Be Reduced 

Worker Would 
Break Down Physically 

Nothing: Worker Wouldn't 
Make More Money 

Would Cause Unemployment 

Other Replies 



JUJ 8% 



Figure 5. Reasons Given by Manual Workers for Withholding Output * 

In a later study, conducted by the Opinion Research Corporation (19) 
for their Public Opinion Index of Industry in 1949, involving a nation- 
wide sample of 1,021 manual workers, 38 per cent expressed the opinion 

* By permission, from "What the factory worker really thinks about productivity, 
nationalization of industry, and labor in politics/' Factory Mgmt. Mainten., (January) 
1946, 104, 80-88. 


that it would be "bad for me" if all workers in the plant turned out more 
work per hour than they do now. This study was made at a time when 
there had been a considerable number of layoffs in American industry. 
Workers were asked if layoffs were likely at their plant in the next three 
months or so. Of those questioned 27 per cent replied "Yes, likely"; 52 
per cent said "No, not likely"; and 21 per cent expressed no opinion. Table 
5 shows that those who fear layoffs are consistently more unfavorable in 
their views on productivity. The table also illustrates further the variety 
of reasons given by manual workers for withholding output. 

TABLE 5. Relation between Fear of Layoff and Attitude towards Increased 

Output (after Public Opinion Index for Industry, Opinion 

Research Corporation, 1949) 

Per cent Replying 

Fear layoffs Do not 

Greater output would be bad for me 50 32 

My company would just add production 

increase to profits 68 46 

Work speed is too fast in my shop 25 16 

Disagree that workers get chief benefit 

of production improvemejxts 64 52 

My company is not doing its best to 

give steady work 38 15 

It is to be noted that seven per cent of the workers questioned in the 
Factory survey referred to above state that the worker would break down 
physically if he turned out more than the average. In this connection, it 
is of interest to speculate upon the functions of voluntary restriction as a 
protective device against overfatigue. Yoder (34, p. 380), for example, 
quotes the observation by Adam Smith that "workmen, when they are 
liberally paid by the piece, are very apt to overwork themselves, and to 
ruin their health and constitution in a few years/' In contrast, Gardner 
points out that while "some theorists have even felt that workers were so 
'money hungry' that they were apt to overwork themselves in piece-work, 
actually such simple faith in the incentive value of piece-work has rarely 
been justified in practice" (7, p. 150). Statements by women workers, 
cited by Davis, such as "What's money compared with losing your 
health?" or "I've come to realize that a rest's worth more than overtime" 
(5, p. 55), may represent expressions on the part of the worker of his 
psycho-physiological limitations in face of the demands of modern in- 
dustry. In addition, the possibility cannot be overlooked that, in some 
instances at least, restriction of output results from a disinclination to 


work, and that this sometimes explains slowdown and perhaps actual 
sabotage on the part of some workers. 

Plant Studies of Output Restriction 

It is apparent that many factors must be considered in an effort to 
explain why workers engage in output restriction. These will be considered 
further in later pages. For the moment, it is sufficient to note that evi- 
dence from employee-attitude surveys clearly supports the opinion that 
unfavorable attitudes toward incentive wages are widespread among 

Both day-to-day observations in industry and studies of voluntary re- 
striction in industry provide additional evidence of voluntary restriction 
of output by employees in unionized and nonunionized plants; in plants 
with, as well as in plants without, wage-incentive systems. There is sound 
ground for the opinion that there is a large reserve of productive poten- 
tial which is not tapped because of voluntary restrictions because of 
"conscientious withdrawal of efficiency" achieved by slowdowns, bun- 
gling, and direct obstruction (28) . It has become evident, as suggested by 
McGregor, that "people often expend more energy in attempting to 
defeat management's objectives than they would in achieving them" 

For many years unions have established a frank policy of controlling 
output in order to increase their power (12a, 13); to stretch out available 
work as a means of curtailing layoffs; to avoid reduction in piece rates, 
and to protect the worker from an undue expenditure of energy ( 16) . Such 
restriction is well exemplified in the limitation of the number of bricks 
by the bricklayers' union; in the refusal of painters' unions to allow the 
use of a brush wider than 4% inches; in the demand by plumbers' unions 
that pipe be cut and threaded on the job, etc. (31 ). Spray guns for paint- 
ing are permitted in some cities, but not in others. "Make-work" prac- 
tices in the building trades, "featherbedding" in the railroad industry, 
represent essentially other, although more indirect ways, of curtailing 
the productivity of individual workers with the avowed purpose of safe- 
guarding job security; of furnishing employment to more union mem- 
bers, or of otherwise protecting the entrenched interests of organized 
labor (2). 



Voluntary limitation of output is not, however, limited to organized 
workers. As Mathewson (17) has shown in an early study, it is practiced 
on an enormous scale by factory and office workers at diverse occupa- 


tional levels in practically every industry. Evidence of this was obtained 
by means of first-hand observations and from interviews with approxi- 
mately 350 workers and 65 executives. While making observations, the 
investigator actually worked as a laborer, machine operator, and in other 
occupations, living with working people in their home environments. 
Interviews included formal conferences with people who had been in- 
formed of the purpose of the inquiry and informal conversations with fel- 
low workers. In addition to observations made by the investigator him- 
self, a group of six workers co-operating in the study made a series of 
separate reports in the form of personal letters from industrial centers. 
Recorded in detail were 223 instances of obvious restriction. These were 
found in 105 establishments in 47 localities, representing 25 classified 
industries and 14 miscellaneous industries. Cases included two types, (a) 
those in which production was actually reduced and (b) those in which 
the intent to restrict was clearly evident, whether or not restriction actu- 
ally resulted. 

The nature of restrictions and the character of the underlying factors 
are well illustrated by data gathered in a machine manufacturing plant, 
where the investigator was one of a gang of three production workers, on 
a group bonus, who believed that if earnings rose above 95 cents an hour, 
the management would reduce rates. Although the men could easily have 
produced approximately twice as much work, output was intentionally 
restricted to the lower figure to prevent rate cuts. In another of the many 
illustrations of output restriction cited by Mathewson, workers on a con- 
veyor line expressed their resentment of "speed-up" by throwing a piece 
of scrap tin into the works, or putting a can in crookedly, in order to stop 
the belt after it had been speeded up 33% per cent by a "cheese engi- 
neer" (17, p. 22). 


Repeated observations in industry confirm Mathewson's conclusion 
that restriction of output, and a strong shop opinion supporting it, exist 
in some of the most enlightened and efficiently organized industries ( 30) . 
As Drucker has pointed out, open restrictions on efficiency and produc- 
tivity, as represented in restrictive union rules and "featherbedding" prac- 
tices, "are only the part of the iceberg that is above the water. Much more 
important are the invisible, unwritten, informal restrictions decreed by 
the custom and common law of every plant" (6, p. 83), as reflected in 
ironclad arrangements to restrict output; a tacit setting of production 
quotas which a worker would be ill-advised to exceed; withholding from 
management of new ways or gadgets developed by workers for doing the 
job better and faster, etc. 


Attempts by management and supervision to overcome such informal 
restrictions may not only strengthen them but even be used by the work- 
ers to enforce their own production code. This is illustrated, according 
to Drucker, in the experience of a large Midwestern candy plant, where 
packers are paid on a sharply rising scale for any days output above the 
standard. Each team of packers manages to turn out a truly phenomenal 
number of filled candy boxes one day each week, thus earning the high- 
est bonus. On all other days each team turns out just enough not to get 
"docked." During the day allotted to a team for record production al- 
ways the same day each week all the other teams in the packing room 
are expected to help the day's high producers! In this way each worker 
manages to earn a good deal more than if he were to exceed his quota 
every day by a small amount, and at the same time this system prevents 
any increase in the production quota. The "bosses" in the plant have 
learned that an attempt to break the production code will bring serious 
trouble, and it is possible, according to Drucker, that they also sympathize 
with the workers' attitude and may themselves share the workers' fear 
that increased productivity will cost some of them their jobs. 


The illustrations presented immediately above represent more or less 
casual observations of restriction, as contrasted with the systematic study 
by Mathewson referred to in earlier pages. A later and comparable study 
by a "social scientist," applying to a single shop, appears in the detailed 
observations and analysis of output restriction by Roy (21), who worked 
for 11 months in 1944 and 1945 as a radial drill operator in the machine 
shop of a steel-processing plant. For 10 months of this period Roy kept 
a daily record of his "feelings, thoughts, experiences, and observations" 
and of conversations with fellow workers. These were noted from memory 
at the end of each workday. Simultaneously, he openly recorded his own 
production in the shop. 

Observations made by Roy led to the conviction that systematic re- 
striction of output was the rule of the shop. Conversations with other 
workers led to the opinion that fear of price cutting was largely at the 
base of restriction in the shop. Workers here firmly believed that if they 
exceeded their quotas where this was possible "the consequence would 
be either reduced earnings for the same amount of effort expended or 
increased effort to maintain the take-home level." 

Jobs in the shop were categorized by workers into two groups, viz., 
"gravy" jobs on which the "bogey" or quota could be attained and 
''stinker 99 jobs where standards were such that the "bogey" could be 


reached only with great difficulty, if at all. Experience in the shops sug- 
gested that "operators, ignoring finer distinctions in job timing, sort jobs 
into two [mental] boxes, one for 'gravy' jobs, the other for 'stinkers' " 
(21, p. 429). Support of this "hypothesis" is found in a bimodal distribu- 
tion of output, observed both in the investigator's own work and in the 
shop as a whole, which suggests, according to Roy, that output restriction 
can be classified into two major types, viz. "quota restriction" and "gold- 
bricking." Quota restriction is a limitation on "gravy" jobs in order not 
to exceed set maximums and thus to avoid cuts in piece rates. Goldbrick- 
ing, by contrast, appears to be a "holding back" or failure to release effort 
when the "bogey" appears to be unattainable. 

During the ten-month diary period, 75 different piece-work jobs were 
assigned to Roy. He reports that he "made out" in only 31 of these. Of 
the 31 jobs, 20 afforded quota earnings of $1.25 an hour or more. By ex- 
tending his efforts past quota limits to find the earning possibilities of the 
jobs, Roy discovered that on 16 of the 20 quota jobs he could have earned 
more than $1.30 per hour. By including four additional jobs on which he 
made "slow starts," and which were not assigned again, Roy found 20 
jobs with excess-quota potentials capable of yielding $1.30 per hour or 

Table 6 presents data on five of the 20 jobs which showed potentialities 
of yielding hourly earnings in excess of $1.30 per hour. Waste time and 
loss in earnings is computed for each job according to maximum earnings 
indicated in each case by actual tests and according to the number of 
hours devoted to each job. "For instance, operation 'pawls/ which leads 
the list with 157.9 total hours worked, showed, by test, possibilities of 
earnings of $1.96 per hour. At potentialities of $1.96 per hour, over 36 
per cent of each hour is wasted when the operator holds his turn-in to 
$1.25 an hour. Total waste time in the 157.9 hours expended on the pawls 
could then be computed at 57.2 hours or over a third of the time actually 
put in. Earnings might have been, at $1.96 per hour, $309.48; whereas, 
at the quota level of $1.25, they would have been but $197.38 a loss of 
$112.10" (21, p. 433). 

Waste time for the 20 jobs was found to total 286 hours, or 36.4 per 
cent of a total 786.5 hours actually put in on them. This represents, on 
the jobs under consideration, a wastage of 2.9 hours on each eight-hour 
day put in, or a total loss of 35.75 days out of 98.3 actually worked. With 
potential earnings of $1,584.43 for the 98 days and with quota earnings 
at $983.18, the wage loss to the worker, Roy points out, would be $601.25, 
or $6.12 per day, or 76 l / 2 cents per hour. "By this logic, if the worker could 
'cut loose' on the 20 jobs listed, he would average $2.01 an hour instead 
of $1.25. And since the 786.5 hours actually put in on the 20 jobs repre- 


seated 58.2 per cent of the 1,350.9 total piecework hours for the period, 
and 42.5 per cent of a grand total of 1,850.5 hours that included all non- 
piecework activity as well, it is evident that losses resulting from quota 
restriction alone could represent wastage of considerable magnitude 
an over-all hourly income loss for 1,850.5 hours of 32^ cents an hour!" 
(21, p. 434). 

TABLE 6. Time and Earnings Losses on Operations with Potentialities of 
Yielding Hourly Earnings in Excess of $130 per Hour (after Roy) 



(per hour) 

(per hour) 

Total waste 
(in hours) 


at $1.25 

Loss in 

NT bases 
Gear parts 








The investigator recognizes that generalization applied to the drill line 
from observation of his own behavior is valid only if it could be estab- 
lished that he was an "average" performer, and that the jobs assigned to 
him were characteristic of those assigned to other operators. Comparison 
of his performance with that of other men in doing the same kind of 
work on the same shift leads him to conclude that he was, in fact, an 
"average" worker. Furthermore, the appraisals made in Table 6 were 
checked by an analysis of his actual production records to determine the 
actual amount of time wasted by him through quota restriction. The 
average time actually wasted through quota restriction for a six-months 
period proved to be 1.39 hours out of every eight. If, Roy points out, his 
wastage of about two hours a day on quota restriction during the last two 
months of employment is accepted as characteristic of the behavior of 
more seasoned operators, efficiency (in the shop) would be at the 75-per- 
cent level, with immediate possibilities for a 33.3 per cent increase in 
production in quota jobs (21, p. 442). 

A summary of his observations particularly with respect to quota 
restriction leads Roy to hypothesize that "if organizational changes 
were instituted to induce operators to abolish quota limits and 'open up' 
production, the writer's discovered maximums would be quickly raised 
to higher levels by the efforts of the group" (21, p. 435) . Similar analyses 
of "goldbricking" practices on "non-make-out piece work" and on "day 
work" are interpreted by Roy as supporting the above generalization and 
the further conclusion that "the concept of 'cultural drag' might be more 


descriptive than 'cultural lag' in depicting the trailing of some of our 
industrial practices behind technological advance" (21, p. 442). 

Exception might well be taken to such generalizations on the ground 
that the observations were limited to a single shop, by a single and possibly 
"biased" observer. However, as shown in Figure 5, a considerable propor- 
tion of workers surveyed by the Opinion Research Corporation for Fac- 
tory were of the opinion that management would raise production quotas 
if the worker turned out more than an average amount. There also ap- 
pears ample evidence from day-to-day observation in industry to support 
the view, as colorfully stated by Hard, below, that the fear of price cut- 
ting has frequently been justified by the experience of the worker. 

"Management sends a time-study man to observe a worker at a ma- 
chine. This lofty character has a watch which divides a minute not 
merely into 60 parts but into a hundred. With its help he decides the 
exact length of time required for a certain operation. 

"So the worker gets to work. He 'speeds up/ He beats the time-study 
man's time. He climbs, let us say, to $1.50 an hour. Then management 
cuts the number of cents per operation till the worker is earning no more 
going fast than he used to earn going slow. 

"This has happened to a million workers in American industry. To 
my knowledge, it has happened repeatedly ever since the war began. It 
makes workers wary. It makes them hold back. It causes great masses 
of them habitually to work W&y below their productive power. Here 
is the greatest single loss of human energy in American life" (11). 

The Decline of "Economic" Man 

In general, studies of the type cited above give continuing support to 
the conclusions reached by Mathewson, as a result of plant observations 
and interviews with management representatives, that 

"1. Restriction is a widespread institution, deeply intrenched in the 
working habits of American laboring people. 

"2. Scientific management has failed to develop that spirit of con- 
fidence between the parties to labor contracts which has been so potent 
in developing good-will between the parties to a sales contract. 

"3. Underwork and restriction are greater problems than over- 
speeding and over-work. The efforts of managers to speed up working 
people have been offset by the ingenuity of the workers in developing 
restrictive practices. 

"4. Managers have been so content with the over-all results of man- 
hour output that only superficial attention has been given to the workers' 
contribution or lack of contribution to the increased yield. Attempts to 
secure increased output have been marked by traditional and unscientific 
methods, while the workers have held to the time-honored practices of 
self-protection which antedate time study, bonus plans, and other de- 
vices to encourage capacity production. 


"5. Regardless of how much the individual may or may not desire 
to contribute a full day's work, his actual experiences often turn him 
away from good working habits" (17, p. 146). 

The existence of antagonism on the part of workers to the concept of 
wage incentives and the operation of wage-incentive plans does not repre- 
sent proof that what is being attempted is essentially wrong (18). 
Neither does employee antagonism represent proof that the employee is 
not motivated by the desire for more money to be used in satisfying basic 
material needs and in raising his standard of living. Nevertheless, it has 
become increasingly evident that the effectiveness of the financial in- 
centive, as embodied in piece-rate and similar systems, can be reduced to 
nil if wants and needs not gratified by the monetary reward are thwarted 
in the formulation or administration of the plan. Group cohesion in the 
restriction of output under wage-incentive plans, such as that observed 
by Mathewson and Roy, and facts uncovered in other studies, have com- 
pletely destroyed the myth of the employee as an "economic man" re- 
sponding with ordered simplicity and as a "rugged individualist" to the 
pressure of material needs. "Certainly," write the Selekmans, "by now 
everyone agrees that a multiplicity of social institutions condition human 
behavior; that the completely rational individual, propelled by self-inter- 
est, making all his decisions in daily living by calculations of incremental 
utility, is a figment of the scientific imagination" (22, p. 114). 

Consideration of the problems of productivity, satisfaction, and morale 
in industry has led many thoughtful observers to the opinion that a 
predominantly economic orientation toward employee motivation, as re- 
flected particularly in the industrial engineer's emphasis upon wage in- 
centives, overlooks the complex nature of human motives (20) . Investiga- 
tions described in Parts III and IV of this volume have made it apparent 
that the desire for money is only one of many motives which influence 
production and determine the satisfaction derived by the worker from 
his job (10). It has become increasingly evident that the worker seeks 
more than a financial reward from his job, and that the extent to which 
other needs and wants are satisfied both has a bearing upon how well 
a wage-incentive system will work in a particular situation, and exercises 
a direct and frequently predominating influence upon employee per- 
formance and satisfaction at work, and upon employee morale. 



1. W. B. D. Brown, "Incentives within the factory/' Occup. PsychoL, 1945, 19, 81- 

2. S. L. H. Burk, Labor Relations and Competitive Costs Hindrances to Full 
Production. Philadelphia: Chamber of Commerce, Industrial Relations Con- 
ference, Feb. 5, 1948, 6-8. 

3. P. A. Cain, Individual Differences in Susceptibility to Monotony (unpublished 
doctoral dissertation). Cornell University, 1942, 58-86. (Cited from T. A. Ryan, 
Work and Effort. The Ronald Press Co., 1947, 199-200). 

4. F. B. Copley, Frederick W. Taylor. Harper and Brothers, 1923, Vol. 1. 

5. N. Davis, "Some psychological effects on women workers of payment by the 
individual bonus method/' Occup. PsychoL, 1944, 18, 53-62. 

6. P. F. Drucker, The New Society. Harper and Brothers, 1950. 

7. B. B. Gardner, Human Relations in Industry. Richard D. Irwin, 1946. 

8. C. S. Golden and H. J. Ruttenberg, "Union participation: key to greater produc- 
tivity," Advanced Mgmt., 1942, 7 (April-June), 55-56. (Reprinted from C. S. 
Golden and H. I. Ruttenberg, Dynamics of Democracy, Harper and Brothers, 

9. W. Gomberg, "The relationship between the union and engineers/' Mech. Eng., 
1943, 65, 425-30. 

10. P. Hall and H. W. Locke, Incentives and Contentment. Sir Isaac Pitman and 
Sons, 1938, Chapter IV. 

11. W. Hard, "Incentive pay: for mdre war production; for more peace prosperity/' 
Readers' Digest, 1943, 43 (August), 11-15. 

12. R. K. Hovel, "Our eleven years with incentives," Factory Mgmt. Mainten., 1948, 

12a. V. D. Kennedy, Union Policy and Incentive Wage Methods. Columbia Univer- 
sity Press, 1945. 

13. C. W. Lytle, "Recent developments in wage incentives," Advanced Mgmt., 1938, 
3 (March), 79-83. 

14. D. McGregor and I. Knickerbocker, "Industrial relations and national defense," 
Personnel, 1941, 18, 1, 49-63. (Reprinted as Department of Economics and 
Social Science, Massachusetts Institute of Technology, Series 2, No. 7.) 

15. D. McKenzie, "Wage incentives," Advanced Mgmt, 1944, 9, 129-35. 

16. C. Madge, "Payment and incentives," Occup. PsychoL, 1948, 22, 39-45. 

17. S. B. Mathewson, Restriction of Output Among Unorganized Workers. The 
Viking Press, 1931. 

17a. D. McGregor, "Changing patterns in human relations," Conference Bd. Mgmt. 
Rec., 1950,12,9, 322 ff. 

18. R. M. Presgrave, The Dynamics of Time Study. McGraw-Hill Book Co., 1945. 

19. "Productivity" from the Worker's Standpoint. Princeton: Opinion Research 
Corporation, 1949. 

20. A Review of Wage Incentive Practice. Melbourne: Department of Labour and 
National Service, 1949. 

21. D. Roy, "Quota restrictions and goldbricking in a machine shop," Amer. /. Soc., 
1952, 57, 427-42. 


22. B. M. and S. K. Selekman, "An economics for administrators/' Harvard Bus. Rev., 
1951, 29, 112-28. 

23. B. M. and S. K. Selekman, "Productivity and collective bargaining," Harvard 
Bus. Rev, 1950, 28, 127-44. 

24. S. H. Slichter, Union Policies and Industrial Management. Washington, D.C.: 
Brookings Institution, 1941. Chapt. 10: "Union attitudes toward basic systems 
of wage payment." Chapt. 11: "Problems and policies created by piecework/' 

25. R. S. Uhrbrock, A Psychologist Looks at Wage Incentive Methods (Institute of 
Management Series No. 15). New York: American Management Assn., 1935. 

26. United States, 62nd Congress, 2nd Session, House of Representatives, Report 
No. 403, Taylor and Other Systems of Shop Management. U.S. Govt. Printing 
Office, 1912. 

27. United States, 63rd Congress, House of Representatives, Report No. 1175, Tay- 
lor System of Shop Management. U.S. Govt. Printing Office, 1914. 

28. T. Veblen, The Engineers and the Price System. The Viking Press, 1921. 

29. Wage Incentives. Princeton: Opinion Research Corporation, 1946. 

30. "Wage incentives restriction of output," Factory Mgnit. Mainten., 1936, 94, 
3, S203-S205. 

31. G. S. Watkins, Labor Problems. Thomas Y. Crowell Co., 1929. 

32. "What the factory worker really thinks about productivity, nationalization of 
industry, and labor in politics," Factory, 1946, 104, 1, 80-88. 

33. "What the factory worker really thinks about his job, unemployment, and in- 
dustry's profit," Factory Mgmt. Mainten., 1947, 105, 12, 86-92. 

34. D. Yoder, Personnel Management and Industrial Relations. Prentice-Hall, Inc., 

Part 2 



Motives, Attitudes, and 

DURING the early 1920's Whiting Williams, fresh from a 
job as Personnel Director and Vice-President of the Hydraulic Steel Com- 
pany, donned overalls and obtained employment in a variety of jobs 
on the railroads, in the mines and factories of this country and Europe 
(82, 84). Underlying this move was an interest in obtaining a first-hand 
view of the attitudes of workers toward their jobs, their associates, their 
managements, their home problems, and toward other groups, condi- 
tions, and activities which might influence their desire to work and their 
interest in work. One conclusion derived from his experience as a work- 
man was that a fundamental factor in man's attitude toward his work is 
the pay check. However, according to Williams, the pressure of the week's 
indispensable pay check is only one of a series of wants for which fulfill- 
ment is sought on the daily job, and the financial incentive represents 
only the beginnings in tapping the resources of the worker. In Main- 
springs of Men, Williams reports a personal experience to illustrate this 
important fact. 

" 'Charley, how'd you like to join the millwright gang?' the foreman 
called to me [Williams reports] . He appeared to think he was offering a 
distinguished honor in spite of his explanation that it paid only 2 cents 
an hour more. The change was accepted with indifference; surely so slight 
an increase in pay could not mean much of a promotion. Half an hour 
sufficed to prove my error. As I came by my former companions, carrying 
oil can and wrench, I made a veritable sensation! Every one of these old 
friends leaned upon his shovel and wiped the sweat and dirt out of his 
eyes while he exclaimed: 

" 'Hey, Boodie! Where you catch-em job? Meelwright gang? Oil can 
and wr-renchl No more . . . shovel! My Ga-wd!' 

"From that moment it was possible to talk familiarly with the first and 



second helpers, those experts who peer through their colored spectacles 
into the changing conditions of the furnace's "bath 9 of 'hot metaV up to 
the instant of the 'tapping. 9 For three weeks I had puzzled why these 
men would have nothing to do with me. Now we were suddenly become 
pdsl But this was not all. My elevation brought honor not only inside 
but outside the plant. Without doubt, if my wife had lived nearby, she 
would have received the congratulations of the wives of the unskilled 
laborers: 'Your man he catch-em fine job! 9 And not one of them but 
would have observed closely, the next day, to see whether she continued 
to speak to theml" (83, p. 56). 

In a series of such dramatic presentations Williams drew forceful atten- 
tion to the significance of the "desire for standing among fellow workers" 
and of other similar wants which the worker brings into the factory or 
which arise in the course of work in the plant. Consideration of the indus- 
trial scene by many thoughtful observers, e.g. Parker ( 54) , Debreuil ( 18) , 
Tead (67), Mayo (45, 46), and the day-to-day experience of management 
with problems of productivity, satisfaction and morale, have also f ocussed 
attention upon the necessity of determining, as accurately as possible, the 
nature and strength of motives which operate in releasing the energies of 


The identification of motives operating in industry or elsewhere, and 
the determination of their strength once they have been identified, pre- 
sent many difficulties. Effective results can be achieved only through 
systematic research conducted within a sound theoretical context. Theory 
is important because it supplies a frame of reference for the integration 
of specific findings, i.e., a framework of analysis which makes it possible 
to comprehend a vast mass of facts and observations (17, 56), and for 
arriving at acceptable generalizations or principles. Furthermore, theory 
is a source of propositions or hypotheses to be tested by further experi- 
mentation in order to reinforce and expand generalizations or principles 
derived from research. 

The purpose of Part II of this volume is to consider briefly the theo- 
retical background of experimental studies and surveys in the area of 
motivation and morale. Familiarity with theory is desirable not alone 
from the viewpoints outlined above, but also because, in the absence of 
exact facts, decisions based upon acceptable theory can be more produc- 
tive than those based upon more questionable theory (5). For the stu- 
dent of industrial psychology the discussion of theoretical background is 
necessary as an aid in the comprehension of underlying problems and in 
the design of research on motivation and morale. The discussion of moti- 
vation and morale in terms of psychological theory can also be of interest 


to readers from industry. On the other hand, nonprofessional readers 
may prefer to bypass some of the more detailed discussion of theory, par- 
ticularly since it involves consideration of terminological distinctions 
sometimes described as the "weird jargon" of the social scientists (80, 81 ) . 
Nevertheless, a consideration of the relations among motives, attitudes, 
and incentives, on pages 76-79, of the "high-lights" of Chapter 5, espe- 
cially pages 83-85 and sections dealing with industrial implications, and 
of the concluding portion of Chapter 6 (pages 1 16-121 ) , is recommended 
to such readers. 

The Instinctive Basis of Economic Activity 

In their search for basic motives in work, a number of psychologists, 
economists and sociologists among them Patten (55), Veblen (74), 
Trotter (73), Tead (67), DeMan (19) pounced upon instinct as the 
prime mover in economic activity. This emphasis on the instinctive and 
presumably impulsive basis of human conduct represented a reaction 
against earlier viewpoints, reflected in the concept of the "economic 
man/' which regarded behavior as wholly rational and volitional in char- 
acter (35). Proponents of instinct as the "prime mover" accepted as a 
fundamental assumption McDougall's early assertions that "instinctive 
impulses determine the ends of all activities and supply the driving power 
by which all mental activities are sustained" * and that "without them 
the human organism would become incapable of activity of any kind; it 
would be inert and motionless like a wonderful clockwork whose main- 
spring had been removed or a steam engine whose fires had been drawn" 
(40, p. 44). 

The significance attached to instincts in industry is perhaps best illus- 
trated in Parker's studies of motives in economic life ( 54) . Man, accord- 
ing to this economist, "is born into this world accompanied by a rich 
psychical disposition which furnishes him ready-made all his motives 
for conduct, all his desires, economic or wasteful, moral or depraved, 
crass or aesthetic" (54, p. 133). Economic, as other forms of activity, is 
motivated by the demand for the realization of these instinctive wants, 
which are the efficient and tried guides of conduct, representing "the re- 
sult of endless experiments of how to fight, to grow, to procreate, under 
the ruthless valuing mechanism of the competition for survival" (54, 
p. 135). 

Every instinct, 2 according to Parker, is subject to the Freudian for- 

1 Not italicized in the original. 

2 The list of instincts used by Parker include sixteen, classified as follows: Instinct 
of gregariousness; instinct of parental bent; instinct of curiosity; instinct of acquisition; 
instinct of fear and flight; instinct of mental activity; the housing or settling instinct; 


mulae of repression, rationalization, and sublimation. If balked, it will 
break out in nervous disorder. If properly directed, its motive force may 
be sublimated into socially beneficial channels (20). If an environment 
interferes with the play of these instincts, the human organism may 
undertake a complete, tenacious, and destructive revolt against it a re- 
volt which, in industry, may express itself in the form of mild restric- 
tion of output, of severe sabotage, or of a rising against the very form of 
organization itself. This revolt expresses not willful, responsible behavior, 
but a stereotyped reaction against these conditions. The forces which 
interfere with instinctive expression in the industrial environment in- 
clude monotonous work, dirty work, specialized work, automatized 
work, the menial role of labor, insecure tenure of the job, injustice in 
hiring and firing, 'seasonal employment, etc. (20, p. 48). Unhealthy 
fixations, irrational inferiority obsessions, are made part of the work- 
er's personality by the presence of these conditions in his daily task. 
Inferiority compensation finds its outstanding expression in the strike, 
the violence of which varies directly with the psychic annoyance, and 
which, in large part, reflects the mental ill-health generated by the 
worker's experience in the plant. The ultimate effect of the interplay of 
these forces is a "true psychosis, a definite mental un-balance, an effi- 
ciency psychosis, as it were,", .which cannot be cured by "our present 
moralizing and guess-solutions" or by "a ten per cent wage increase" (20, 
p. 50). 

The defence of the instinct hypothesis by Parker rests on a study 
of activities and conditions of work of migratory workers the casual 
laborers of the western and northwestern parts of the United States. 
However, according to this author, conflict appears wherever industrial 
conditions fail to satisfy the instinctive wants. 


In contrast to the popularity, not so many years ago, of the instinct 
hypothesis, most current discussions of motivation relegate instinct to a 
relatively unimportant position in the consideration of human motiva- 
tion. The questions to be answered in deciding whether behavior can be 
regarded as instinctive center around the extent to which an organism 
exhibits behavior which is unlearned (68), and the identification of those 
particular aspects of its behavior which are unlearned (32) . Examples of 
complex behavior which the organism has had no opportunity to learn 
can easily be found among lower animals, insects (55a), birds (58a), 

instinct of migration; instinct of hunting; instinct of anger; instinct of revolt at con- 
finement; instinct of revulsion; instinct of leadership and mastery; instinct of sub- 
ordination; instinct of display; instinct of sex. 


fishes (68a) , and some mammals, e.g., rats (2a) . However, the role played 
by environmental determiners in the organization of human behavior, 
even in connection with biological processes, is so important that it is very 
difficult, if not impossible, to point to instances of human behavior which 
meet the criteria of instinct. 

Hayes (29), for example, points out that the use of the concept of 
instinct as an explanation for human behavior is particularly hard to 
defend when behavior has relevance to social situations. Illustrations of 
the systematic variation of human motives and attitudes under different 
cultural patterns can be found in the work of anthropologists, such as 
Benedict (2b). The weight of anthropological evidence on the occur- 
rence of competition, to select but one example from many, leads to the 
conclusion that there is by no means a "competitive instinct" among 
humans as James, for instance, would have us believe when he speaks 
of "emulation or rivalry" ( 31, II, p. 408) . The same may be said to be true 
of the other social behaviors which have from time to time been included 
in lists of instincts postulated by James (31), McDougall (40), Parker 
(54), Tead (67), and others. 

Varieties of Motives 

It is now generally acknowledged that the tendency of earlier writers 
to regard any behavior as "explained" once it had been labeled "instinct/' 
represented an oversimplification of the problem of human motivation 
both in industry (79) and in other aspects of daily life (47). As a re- 
sult, attention has been directed toward a broader systematic approach 
(32) which regards adult motives as infinitely varied and complex (2), 
and as derived from the operation not alone of an internal system, but 
from the interaction of the individual and his environment. 

In the development of current views of motivation, the many instincts 
of James (31) and McDougall (40) were questioned and also reduced 
in number by Watson (77) . An even more radical position was taken by 
Kuo (37a) who suggested that the concept of instinct be done away 
with entirely in discussing the behavior of any species. However, al- 
though the work of Cannon (8), Coghill (15), Loeb (38)> Richter (58), 
Sherrington (61), Stone (65), and others has contributed to a more 
detailed analysis of biological mechanisms in behavior, it has not led 
to the abolition of the concept of instinct. Many psychologists con- 
cerned with motivation make use of the concept, but also take the posi- 
tion that instincts belong to a broader class of behavior, associated with 
the operation of biological mechanisms, commonly referred to as pri- 
mary motives. The multiplicity of complex human motives allegedly 
derived from the latter by processes of differentiation, integration, condi- 


tioning, sublimation, substitution, etc., as the result of the interaction 
between organism and environment, are generally classified as secondary 
motives. While this may be considered an oversimplification of the situa- 
tion, it nevertheless appears to represent an accurate condensation of one 
major trend in the current approach by psychologists to the problem of 


One outcome of this trend has been the emergence of classifications 
of dynamic mechanisms designated by the terms need, drive, and mo- 
tive. 41 Although these are frequently used interchangeably in discussions 
of motivation, fairly sharp and systematic distinctions have been made 
in the case of the sb-called primary or biological mechanisms. 

The basis for the term need, in the consideration of such mechanisms, 
lies, in general, in the fact that organisms are not closed systems, com- 
pletely independent of their environment. They expend energy and to 
survive they must maintain in their bodies a supply of the materials from 
which that energy is derived. Even more critical for the continuation 
of an organism's life are the many physical and chemical characteristics 
of its tissues which must be maintained within very narrow limits, e.g., 
the acidity of the blood, the .body temperature, the water content of 
the blood, the oxygen supply of the blood and other tissues, to mention 
a few. An organism which momentarily lacks any of the material necessary 
for the maintenance of life is said to be in a state of need. Need, there- 
fore, is used rather precisely, in a physiological sense, to designate states 
of bodily deficit or the approach of physical and chemical conditions to 
those limits beyond which life is not possible (47). It is, however, possi- 
ble also to conceive of real or imagined deprivations in a purely psycho- 
logical context where no question of organic survival is raised. Such ana- 
logues of the physiological deficits are also called needs 5 (47, p. 437) . 

Many conditions of need are accompanied by activity or behavior on 
the part of the organism. The correlation between need and activity is, 
in some cases, so marked that the behavior of laboratory animals can 
be controlled by withholding food or water and thus artificially inducing 

* Variations of the basic pattern include such classifications as Woodworth's trans- 
formation of "mechanisms' into "drives" (85), Morgan's "biogenic" (or "viscero- 
genic") and "psychogenic" needs (47), Tolman's "superordinate" and "subordinate" 
goals (71), Lewin's "true needs" and "quasi-needs" (39), Stern's assumption with 
respect to the conversion of "phenomotives" into "genomotives" (63), Allport's 
concept of "functional autonomy" (2), etc. 

* Along with these there is frequent reference to 'wants, wishes, and desires. 

* The distinction here is between what are frequently designated as biogenic and 
psychogenic needs, respectively. 


a need state. It is possible, in general, to distinguish two aspects of ac- 
tivity related to needs. One aspect is related to the fact that organisms 
show a simple increase in the number and vigor of their movements or, 
as some prefer to say, they become restless or act randomly, in an undi- 
rected fashion. This phase of need-related activity is said to give evidence 
of drive, since it seeins to represent the release of organic energy without 
any particular direction or guidance (25, 30) . 

Under normal conditions this undirected behavior gives way to be- 
havior which can be viewed as oriented toward the achievement of some 
end state or god. Characteristic of this directed behavior is the fact that 
the organism does not behave randomly, but in a more or less persistent 
fashion performs a series of acts appropriate to the achievement of the 
goal. Within the bounds of such a theoretical approach, the term motive 
is used to describe any condition of the organism which permits the ap- 
pearance of such directed behavior (59, 66). This definition makes it 
possible to designate behavior as motivated even in situations where it 
is not possible to identify corresponding needs. It is important to note 
that the condition referred to as a motive is not simply the need existing 
or presumed to exist at the time the motivated behavior occurs. Also 
necessary for the appearance of directed behavior are conditions facilitat- 
ing muscle co-ordination, sensory discrimination and, in some cases, the 
ability to solve problems and to remember solutions accomplished ear- 
lier (71). 

The richness and complexity of human secondary or "derived" mo- 
tives, together with current inability to identify the corresponding needs 
with the precision possible in the case of physiological needs, produce an 
atmosphere in which motivational terms are often used so loosely as to 
reflect little more than the personal preferences of various writers (66). 
Thus, in discussion of complex motivated behavior in social situations, 
one psychologist speaks of a need for recognition (49, p. 145); another 
of a drive for recognition (57, p. 275), while a third discusses striving for 
status in a context which suggests that a status motive exists (27, p. 168) . 
In succeeding sections of this volume, the terminology favored by each 
investigator will be used in reporting his findings and point of view with 
respect to the nature and sources of motivation. The reader will also find 
instances in which these terms have been used by the author as a cluster 
of concepts in the interpretation of findings from diverse studies. 


There is considerable agreement among psychologists in categorizing 
motives as primary or secondary in character, but there are many de- 
tailed variations among classifications of motives within each of these 


major categories. Systematic classifications are to be found in the litera- 
ture on motivation, including publications by Maslow (44), Murray 
(49), Tolman (71), Morgan (47), and others. A few illustrations, taken 
from such classifications, are cited below. 

It is generally agreed, as stated above, that the primary motives are 
those which serve some biological function for organisms or species. Com- 
monly included in classifications of primary motives are those involved 
in obtaining nourishment and those involving reproduction. Representa- 
tive of other motives classified as primary are elimination, sensory ex- 
perience, and exploration. 

There is general agreement that secondary motives are largely social 
in nature. Frequently mentioned as belonging to this class are motives to 
achieve status (or- some equivalent); to affiliate with other individuals 
or groups, and to achieve some measure of self-esteem. 

Several psychologists, e.g. Newcomb (52), have subdivided the sec- 
ondary motives so as to produce a third class, referred to as personal mo- 
tives. Personal motives are regarded as generally social with respect to 
their derivation, but are distinguished from motives common to all the 
members of a group in ttiat they are involved in the pursuit or main- 
tenance of standards, and. ideals which are "private" and peculiar 
to the individual' art& riqjt sbbred with most other members of the 


There is no right (in the sense of universally accepted) classification 
of motives, although there is some similarity in the patterning of classi- 
fications favored by individual psychologists. Moreover, as stated earlier, 
classifications of human motives are generally characterized by an em- 
phasis upon the role of secondary or "social" motives in the behavior 
of the individual of motives which arise out of or are structured by 
the individual's identification with one or another group (48, 69). Fur- 
thermore, the position has been taken that there are differences in the 
order of strength or prepotency of motives (44, 49). In some instances, 
the order is regarded as independent of the classification of motives as 
primary or secondary. In others, as in the case of the classification of 
"needs" by Maslow, these are said to arrange themselves in a hierarchy 
of prepotency, from the most elemental (physiological needs) to those 
(higher) needs which underlie the higher development of the individual 
(44). According to Maslow, "the most prepotent goal will monopolize 
the consciousness and will tend of itself to organize the recruitment of 
the various capacities of the organism. The less prepotent needs are 
minimized, even forgotten or denied. But when a need is fairly well satis- 
fied, the next prepotent ('higher') need emerges, in turn to dominate 


the conscious life and to serve as the center of organization of behavior, 
since gratified needs are not active motivators" (44, p. 394). 


In dealing with problems of motivation in industry there is wide ac- 
ceptance of the view inherent in a number of motivational theories 
that seeking or avoidance (i.e., the action, the behavior) is associated 
with the existence of a tension system, or state of disequilibrium, which 
the organism seeks to restore to a state of rest or balance. In this sense, 
the terms "tension system" and "condition of disequilibrium" usually 
refer to the conditions previously referred to as needs, while "the terms 
'motive' and 'motivation' are ... used to refer to the inner directing 
processes which determine the movement or behavior towards ends or 
goals. They represent conditions of tensions or disequilibrium within the 
person, with the ensuing behavior serving in some measure to relieve the 
tension or the disturbed equilibrium" * (34, p. 217). 

The Nature and Role of Attitudes 

As suggested earlier, a major outcome of recent developments in moti- 
vational theory presented in this chapter has been to focus attention upon 
a large variety of motives, drives, and needs which are taken to represent 
the mainsprings of human behavior. It is increasingly recognized that 
the solution of problems of production and morale in industry involves 
the close consideration of the wants of the worker which reflect either 
previously established tension systems in the individual or the effect upon 
him of the immediate social situation. 

The possibility of identifying such wants is complicated by the fact 
that motivation as such or more specifically motives, drives, and needs 
cannot be observed directly. Experimental inquiries have disclosed tis- 
sues, glandular mechanisms, and hormones which are involved in moti- 
vated behavior (3, p. 36) . In general, however, it is possible only to infer 
the existence of drives, needs, and wants (34, p. 217), in part, from ob- 
served changes in behavior, especially in controlled experimental situa- 
tions; in part, from measurements of attitudes which express the way in 
which and the extent to which given objects or situations are felt to satisfy 
wants, needs, desires, etc. 


As in the case of other psychological concepts, there is far from uni- 
form agreement in defining the term attitudes and in the delineation 

6 Generally excepted from such treatment are 'lower-order" activities, such as 
reflex response, directly referrable to specific bodily mechanisms "triggered" by par- 
ticular environmental stimuli without the intervention of motives, drives, or needs. 


of psychological processes involved in the formation and expression of 
attitudes. 7 Common to most definitions is the view that an attitude is a 
fundamental state of readiness for motive arousal or a reaction in a char- 
acteristic way to certain stimuli or stimulus situations (52, 62). Further- 
more, it is commonly assumed that attitudes are charged with affective 
or value properties in varying degrees (60). Thus, according to Korn- 
hauser, "An attitude is a more or less specific tendency or readiness to 
react toward (or away from) an object, person, or relation; it is a particu- 
larized motivation or 'set/ usually with marked favorable or unfavorable 
feelings toward the object of the attitude" ( 34, p. 218) . 

From this point of view, attitudes may be "conceived as integrations 
mediating between the fundamental psychological processes and reac- 
tions" (37, p. 131); As a result, psychologists have used the measurement 
of attitudes as a device both for making inferences concerning motives 
and as predictors of behavior. It is from this point of view, for example, 
that the measurement of attitude toward layoff has been viewed as pro- 
viding information concerning the need for security. Likewise, the deter- 
mination, by means of attitude surveys, that attitudes toward wages are 
unsatisfactory has been interpreted to mean that a change in wage rate 
will stimulate basic drives leading to desired conduct in the way of higher 

There has, of necessity, been much oversimplification both in making 
inferences concerning motives and in prediction from the measurement 
of attitudes. Furthermore, as will be made clear in Chapter 12, there is 
reason for skepticism concerning the effectiveness of available attitude 
measures as predictors of behavior. Nevertheless, both verbal and be- 
havioral measures of attitudes have yielded valuable information concern- 
ing the "feeling attachments" or "affective properties" of attitudes and 
have thereby provided a basis for a more complete understanding of what 
workers consider to be important or unimportant in the specific condi- 
tions, policies and practices of the industrial plant. 

The fact that attitudes, motives and conduct are closely interrelated 
has suggested the possibility that a change in attitudes can induce a 
change in behavior, by influencing the direction of motivation or through 
actual modification of "the organized pattern of motivation." An orienta- 
tion toward the modification of attitudes as a way of influencing behavior 
is found in Allport's early view that the concept of attitude "is probably 
the most descriptive and indispensable concept in contemporary psy- 
chology" (1) and in the definition, by Bogardus (4) and Folsom (22), 

7 Contemporary definitions of attitude are to be found in Allport (1, 2), Cattell 
(11, 12, 13), Newcomb (52), Sherif and Cantril (60), and Krech and Crutchfield 


of social psychology as the scientific study of attitudes (21). The signifi- 
cance of attitudes in relation to social conduct is also seen in more recent 
approaches to the study of industrial and social behavior. Thus, according 
to Sherif and Cantril, "the basic psychology of ego formation 8 is essen- 
tially the psychology of the formation of attitudes" (60, p. 94) . Similarly, 
from the viewpoint of the "perceptual" approach 9 in the consideration 
of psychological processes, changes in attitudes associated with differences 
in the perception of a social situation are viewed as having a distinct bear- 
ing upon purposive behavior (9, 28). Furthermore, the application of 
field theory 10 to the reactions of individual members of groups is, accord- 
ing to Brown, "concerned with the effect of field structure on the indi- 
viduals' attitudes toward the groups in which they have membership- 
character and toward other groups" (6, p. 97). 

Presentation of the above view of the relation between changes in at- 
titudes and conduct must not be viewed as suggesting that modification 
of attitudes in a desired direction is invariably or even generally ac- 
companied by a change of behavior in the desired direction. Thus, for 
example, the development of more favorable attitudes toward Negroes 
on the part of an employer will not necessarily be followed by an immedi- 
ate change in policy with respect to hiring Negroes. It is not to be ex- 
pected that the improvement of employee attitudes toward a wage-incen- 
tive plan will invariably be followed by a stimulation of motives to the 
point where the employee "works harder" in order to turn out more 

In some instances, such transition to conduct is not accomplished be- 
cause the "state of readiness" or "susceptibility to stimulation capable 
of arousing the motive" is not sufficiently strong to compete with other 
attitudes associated with other objects, e.g., the small working group and 
its production norms, the union and its policies, etc. Other factors help 
to account for the absence of a one-to-one relationship between attitude 
and action. In fact, a pressing problem in both theoretical and applied 
psychology is the arrival at a better understanding of the mechanisms 
through which the transition from attitude to performance can be facili- 
tated. However, as will be seen from later chapters, evidence exists that 
this can be accomplished in industry, as in other situations. Furthermore, 
even when behavior is not changed, satisfaction and adjustment can be 
increased through programs designed to mould or modify attitudes be- 
cause of the effects of such programs upon the affective aspects of atti- 
tude (50). 

8 See pages 98-10 
See pages 93-98 
1 See Chapter 6. 


The Nature and Function of Incentives 

A major objective in the direction and control of human behavior is 
to influence attitudes and release, direct, or control motives which will 
result in desired activities. As Bird has suggested, "we must have knowl- 
edge of the forms of social control which will release and facilitate de- 
sired activities and which will reduce behavior likely to handicap success 
in cooperative activities" (3, p. 61). Specifically, the problem, from the 
viewpoint of management, is to release those psychological forces which 
will lead the employee to participate willingly, fully and satisfyingly in 
furthering the production aims and other goals of the industrial organiza- 

The conditions which start or initiate, decrease or speed up, or par- 
tially inhibit and direct activities are commonly known as incentives. 
From this point of view, incentives are situations which function in arous- 
ing dynamic forces in the individual (43, p. 97), or arrangements of con- 
ditions introduced with the expectation of influencing or altering the be- 
havior of people. Thus, according to Warren 

"If the motivation of conduct be conceived as originating in some 
maladjustment between orgapism and environment, the organism's 
efforts are said to be the result of inner drive while the critical environ- 
mental conditions are called the incentive; i.e., food is an incentive to a 
hunger driven animal or person; a badge of honor is an incentive to one 
under the urge of ambition" (76, p. 134). 

Incentives may be positive in character, in the sense of facilitating or 
promoting a particular form of behavior, or negative in the sense of in- 
hibiting or hampering response of one kind or another. Whether positive, 
such as material rewards, praise, anticipated success, etc., or negative, such 
as reproof, penalties, removal of privileges, etc., "incentives do not cause 
or initiate die behavior to which they are directed. . . . The immediate 
function of the incentive is to tap motives and to change attitudes which 
in turn redirect effort" ( 3, p. 61 ) and modify behavior. 

Applying this viewpoint, praise (perceived as appropriate and sincere) 
functions as a positive incentive to the individual and, through the grati- 
fication of basic motives (possibly the need for status or recognition) 
and the stimulation of favorable attitudes, releases energy or directs be- 
havior towards improved performance in the way of increased production 
with accompanying feelings of satisfaction and relief from tension. Typi- 
cal relations which may be postulated under sucljan^proach, as adapted 
from Charron ( 14) , are as follows 




Economic security 
Emotional or personal 

Recognition status 


Higher production 
Fewer absences 
Decreased lateness 
Improved quality of 

Active participation in 

company programs 


Fair wages 

Pension plan 

Compensation for sick- 
ness and disability 

Death benefits 


Profit sharing 

Good working condi- 

Pleasant relations with 


Knowledge of results 

Development of skills 

Recognition of efforts, 

Both research findings and practical experience have shown that men 
"exert great effort, endure hardships, overcome obstacles" only when 
"strong motives impel the men concerned toward a goal they desire" (62) . 
Measurements of employee attitudes, which express the way in which 
and the extent to which given objects or situations are felt to satisfy per- 
sonal wants or relieve tension, can contribute to a greater understanding 
of the effectiveness of incentives in releasing energy and directing motives 
toward the achievement of desirable goals. 

The psychological mechanisms through which incentives operate to 
produce such results are far from completely understood. However, nu- 
merous laboratory experiments on animals and children suggest that 
previously "neutral" objects such as poker chips or brass discs can 
come to serve as incentives in that they will be worked for, begged for, and 
subjects will actively learn in order to obtain them. This "incentive" char- 
acteristic appears most clearly when the "neutral" object is made relevant 
to some goal which is important for the individual. Thus Wolfe (86) 
trained chimpanzees to work for poker chips which they could later ex- 
change for food. Wolfe's animals also begged the poker chips from each 
other and from the experimenter. 11 

From a consideration of laboratory and industrial studies, Mace has 
concluded that an incentive functions by engendering a specific inten- 
tion commonly, in industry, the intention to perform a determinate set 
of movements resulting in some industrial product conforming to certain 
standards. According to this viewpoint, "Supervision, verbal encourage- 
ment or reproof, the prescription of standards and so forth, are of value, 

11 These and other investigations arc well summarized by Miller (46a). 


just in so far as, directly or indirectly, they control the specific intention 
which is operative in the performance of the given task. They may be 
useless or even harmful in releasing energy in ways which do not con- 
tribute to the performance of that particular task" (42, p. 2). 

In presenting this viewpoint, Mace further suggests that an intention 
varies (1) in direction, i.e., in what, precisely, the individual is trying to 
do, or in how 'Tie conceives his task"; (2) in intensity, i.e., in how strongly 
he feels impelled to perform it, and (3) in duration, i.e., in the time dur- 
ing which the individual maintains effective concentration. According 
to this point of view, the study of incentives in industry involves, to a large 
extent, a determination of the situations or arrangement of conditions 
which are most effective in stimulating the direction, intensity, and dura- 
tion of that intention which arouses attitudes and produces behavior 
which are considered most advantageous from the point of view of achiev- 
ing industrial goals. While the concept of intention, as used by Mace, is 
open to question in terms of current dynamic theories of motivation, pri- 
marily because of the emphasis upon conscious and volitional processes, 
it is nevertheless of interest as representing a frame of reference which 
has been extensively used, particularly by British investigators, in studies 
of incentives in industry. 

Motives, Attitudes, and Incentives in Industry 

A consideration of current trends in theories of motivation reveals a 
number of changes which have particular significance in the study of 
workers' motives and in the evaluation of incentives. In the first place, 
the viewpoint that wages represent the sole incentive to employee produc- 
tion and co-operation, and the single or even primary source of satisfac- 
tion at work, runs counter to motivational theory derived from laboratory 
research and substantiated in industrial studies described in succeeding 
chapters (79). A major outcome of recent developments in systematic 
psychology has been to discredit the major psychological assumption 
made both by business men and economists subscribing to classical eco- 
nomic orthodoxy, viz., that "production and exchange are predominantly 
(if not exclusively) motivated by the attempt to maximize money gains" 
(29, p. 289) . In addition, the effort on the part of economists (supported 
by psychologists) to broaden the psychological base for the exploration 
of economic behavior by reference to instincts finds little or no support 
in current motivational theory. 

On the positive side, current theory postulates a large variety of motives, 
drives, needs, and wants and recognizes the significance of attitudes and 
sentiments as overt expressions of the extent to which internal or external 
conditions, including incentives, can function in stimulating behavior 


appropriate to the relief of tension and to the restoration of equilibrium 
within the person, with ensuing satisfaction to the individual. In this 
respect, current motivational theory directs attention to the study of 
motives which were overlooked or disregarded in the earlier and inade- 
quate classical psychological generalization of economics. Of particular 
importance, as will appear more clearly in the following chapters, is the 
recognition of the extreme significance of environmental (cultural, so- 
cial) factors in the integration of motives, drives, and needs, leading to 
an organic conception of industry as a social institution and the further 
elaboration of a conception of society in which economic activities take 
their place as one aspect of the whole social process (78). Of equal sig- 
nificance is the increasing tendency to approach the problem of motiva- 
tion in terms of the person as a whole in terms of a total personality 
functioning in a natural social environment (10). 

In general, as suggested earlier, the concepts of current motivational 
theory provide an indispensable framework for systematic research on 
motives in industry and on the value of particular incentives. However, 
the solution to the problem of motivation (as to other psychological 
problems) is to be found not in theory, but in systematically designed 
experimental studies conducted, in so far as possible, in the industrial 
plant. Such research, involving the use of industrial workers as subjects, 
presents many technical and practical difficulties (26). Whatever the 
difficulties, "the problems are so central and inescapable and so clearly 
problems of industrial psychology if it chooses to be a social science 
as well as a managerial technique that it appears well worth our while 
canvassing the possibilities of further promising research in the field. 
Among the few greatest questions of our age is that which asks what mod- 
ern industry means to the individual worker with reference to his satis- 
factions and fullness of life" (36, p. 351). The accumulation of evidence 
and the formulation of adequate scientific explanations of the factors 
basic to these represent a continuing challenge to social scientists and 
management alike. 


1. G. W. Allport, "Attitudes" in C. Murchison (Ed.), A Handbook of Socid Psy- 
chology. Clark University Press, 1935, 798-844. 

2. G. W. Allport, Personality: A Psychological Interpretation. Henry Holt & Co., 
1937, 194 ff. 

2a. F. A. Beach, "Analysis of the stimuli adequate to elicit mating behavior in the 

sexually inexperienced male rat," /. Com/). Psychol. P 1942, 33, 163-207* 
2b. R. Benedict, Patterns of Culture, Houghton Mifflin Co., 1934. 


3. C. Bird, Social Psychology. D. Appleton-Century Co., 1940. 

4. E. Bogardus, Fundamentals of Social Psychology. Century, 1931. 

5. S. H. Britt, Social Psychology of Modern Life. Farrar and Rinehart, 1941, 78-79. 

6. J. F. Brown, Psychology and the Social Order. McGraw-Hill Book Co., 1936. 

7. C. Buhler, "Das Problem des Instinktes," Z. fur Psychol., 1927, 103, 46-64. 

8. W. B. Cannon, The Wisdom of the Body. W. W. Norton & Company, Inc., 1932. 

9. H. Cantril, "The Nature of social perception," Trans. N.Y. Acad. Sci., 1948, 10, 

10. H. Cantril, The "Why" of Man's Experience. The Macmillan Co., 1950. 

11. R. B, Cattell, The Description and Measurement of Personality. World Book 
Co., 1946. 

12. R. B. Cattell, "The ergic theory of attitude measurement," Educ. and Psychol. 
Meas. 9 1947, 7, 221-46. 

13. R. B. Cattell, E. F. Maxwell, P. H. Light, and M. P. Unger, "The objective 
measurement of attitudes," Brit. /. of Psychol. (Gen'l Section), 1949, 40, Part 2, 

14. L. Charron, "Increasing employees' job interest/' Proceedings: The Fifth Annual 
Training Conference of Educational Directors in Industry and Commerce. Mon- 
treal, Canada: Canadian Industrial Trainers Assn., 1951, 53-62. 

15. G. E. Coghill, Anatomy and the Problem of Behavior. Cambridge University 
Press, 1929. 

16. G. E. Coghill, "The structural basis of the integration of behavior," Proc. Nat. 
Acad. Sci., 1930, 16, 637-43. 

17. E. Dale, "Company research in human relations," AMA Association News, 
June 23, 1949, 22, 3-4. 

18. P. C. Debreuil, Robots or Men? Harper and Brothers, 1930. 

19. H. DeMan, Joy in Work. Henry Holt Co., 1929. 

20. Z. C. Dickinson, Economic Motives. Harvard University Press, 1922. 

21. E. Paris, "Are instincts data or hypotheses?," Amer. /. Soc., 1921, 27, 193-94. 

22. J. K. Folsom, Social Psychology. Harper and Brothers, 1931. 

23. J. C. Flugel, Man, Morale and Society. International Universities Press, Inc., 1945. 

24. E. R. Grove, "Sociology and psychoanalytic psychology," Amer. J. Sci., 1917, 23, 

25. E. R. Guthrie, The Psychology of Human Conflict. Harper and Brothers, 1938. 

26. P. Hall and H. W. Locke, Incentives and Contentment. Sir Isaac Pitman and 
Sons, 1938. 

27. C. M. Harsh and H. G. Schrickel, Personality: Development and Assessment. The 
Ronald Press Co., 1950. 

28. A. H. Hastorf and A. L. Knutson, "Motivation, perception, and attitude change," 
Psychol Rev., 1949, 56, 88-94. 

29. S. P. Hayes, Jr., "Some psychological problems of economics," Psychol. Bull., 
1950, 47, 289-330. 

30. E. B. Holt, Animal Drive and the Learning Process. Henry Holt & Co., 1931. 

31. W. James, Principles of Psychology. Henry Holt & Co., 1890. 

32. F. W. Irwin, "Motivation" in H. Helson (Ed.), Theoretical Foundations of 
Psychology. D. Van Nostrand Co., Inc., 1951, 200-47. 

33. Z. Y. Kuo, "The net result of the anti-heredity movement in psychology," Psychol. 
Rev., 1929, 36, 181-99. 

34. A. W. Kornhauser, "Analysis of 'class' structure of contemporary American so- 


ciety" in C. W. Hartmann and T. M. Newcomb (Eds.), Industrial Conflict. The 
Cordon Co., 1939, 199-264. 

35. A. W. Kornhauser, "The motives-in-industry problem/' Ann. Amer. Acad. Pol. 
Soc. Sci., 1923, 110, 105-16. 

36. A. W. Kornhauser, "The study of work feelings," Pers. /, 1930, 8, 348-51. 

37. D. Krech and R. S. Crutchfield, Theory and Problems of Social Psychology. 
McGraw-Hill Book Co., 1948. 

37a. Z. Y. Kuo, "A psychology without heredity," Psychol. Rev., 1924, 31, 427-448. 

38. J. Loeb, Forced Movements, Tropisms, and Animal Conduct. ]. B. Lippincott 
Co, 1918. 

39. K. Lewin, "Behavior and development as a function of the total situation" 
in L. Carmichael (Ed.), Manual of Child Psychology. John Wiley & Sons, Inc., 

40. W. McDougall, Introduction to Social Psychology. Methuen & Co, Ltd, 1908. 

41. W. McDougall, Psycho-analysis and Social Psychology. Methuen & Co, Ltd, 

42. C. A. Mace, Incentives, Some Experimental Studies. (Industr. Health Res. Bd. 
Report 72). H. M. Stationery Office, 1935. 

43. N. R. F. Maier, Frustration. McGraw-Hill Book Co, 1949. 

44. A. H. Maslow, "A theory of human motivation," Psychol. Rev., 1943, 50, 370- 

45. E. Mayo, The Human Problems of an Industrial Civilization. Harvard Business 
School, Division of Research, 1946. (First printing by The Macmillan Co, 1933.) 

46. E. Mayo, The Social Problems of an Industrial Civilization. Harvard University 
Press, 1945. 

46a. N. E. Miller, "Learnable drives and rewards," in S. S. Stevens (Ed.), Handbook 
of Experimental Psychology. John Wiley & Sons, Inc., 1951 (Chapter 13). 

47. C. T. Morgan, Physiological Psychology. McCraw-Hill Book Co, 1943. 

48. G. Murphy, Personality, A Biosocial Approach to Origins and Structure. Harper 
and Brothers, 1947. 

49. H. A. Murray, Explorations in Personality, Oxford University Press, 1938. 

50. P. H. Musscn, "Some personality and social factors related to changes in chil- 
dren's attitudes toward Negroes/' /. Abn. Soc. Psychol, 1950, 45, 425-51. 

51. E. Nelson, "Attitudes: I. Their nature and development," /. Gen. Psychol, 1939, 
21, 367-99. 

52. T. M. Newcomb, Social Psychology. The Dryden Press, Inc., 1950. 

53. OSS Assessment Staff, Assessment of Men. Rinehart and Co, Inc., 1948. 

54. C. Parker, The Casual Laborer and Other Essays. Harcourt, Brace, and Howe, 

55. S. Patten, 'The mechanism of mind," Ann. Amer. Acad. Pol. Soc. Sci, 1917, 71, 

55a. G. W. Peckham and E. G. Peckham, Wasps, Social and Solitary. Houghton 
Mifflin Co, 1905. 

56. M. G. Preston, "Methodological considerations" in H. Helson (Ed.), Theoretical 
Foundations of Psychology. D. Van Nostrand Co, 1951 (Chapter 1). 

57. Quoted from an unidentified source by M. Sherif and H. Cantril, The Psychology 
of Ego-Involvements. New York: John Wiley and Sons, Inc., 1947, 275 f. 

58. C. P. Richter, "Behavior and endocrine regulators of the internal environment," 
Endocrinology, 1941, 28, 193-195. 

58a. W. Rowan, The Riddle of Migration. The Williams & Wilkins Co, 1931. 


59. L. F. Shaffer, The Psychology of Adjustment. Houghton Mifflin Co., 1936, 

60. M. Sherif and H. Cantrfl, The Psychology of Ego-Involvements. John Wiley & 
Sons, Inc., 1947. 

61. C. S. Sherrington, The Integrative Action of the Nervous System. Constable, Ltd., 

62. R. Stagner, "Psychological aspects of industrial conflict II. Motivation," Pers. 
Psychol^ 1950, 3, 1-15. 

63. W. Stern, Allgemeine Psychology auf Persondistischer Grundlage. The Hague: 
Nijhoff, 1935. 

64. C. P. Stone, "Recent contributions to the experimental literature on native or 
congenital behavior/' Psychol. Bull, 1927, 24, 35-61. 

65. C. P. Stone, "Sex drive" in E. Allen, C. H. Danford and E. A. Daisy (Eds.), Sex 
and Internal Secretions. The Williams and Wilkins Co., 1939. 

66. P. M. Symonds, The Dynamics of Human Adjustment. D. Appleton-Century 
Co., 1946. 

67. O. Tead, Instincts in Industry. McGraw-Hill Book Co., 1918. 

68. N. Tinbergen, The Study of Instinct. Clarendon Press, 1951. 

68a. N. Tinbergen, "An Objectivistic study of the innate behavior of animals," 
BibUoOt. Biotheor. 1942, 1, 39-98. 

69. E. C. Tolman, "Physiology, psychology and sociology" (Psychol. Rev., 1938), 
reprinted in E. C. Tolman, Collected Papers in Psychology. University of Cali- 
fornia Press, 1951, 179-89. 

70. E. C. Tolman, "Psychological man" (J. Soc. Psychol., 1941), reprinted in E. C. 
Tolman, Collected Papers in Psychology. University of California Press, 1951, 
207-18. - -/ 

71. E. C. Tolman, Purposive Behavior in Animals and Men. Century Co., 1932. 

72. L. T. Troland, The Fundamentals of Human Motivation. D. Van Nostrand Co., 

73. W. Trotter, Instincts of the Herd in Peace and War. New York: The Macmfllan 
Co., 1919 (2nd Ed.). 

74. T. Veblen, The Instinct of Workmanship. The Viking Press, 1918. 

75. M, S. Viteles, "Man and machine relationship" in R. B. Ross (Ed.), Proceedings 
of 1950 Annual Fall Conference, Society for Advancement of Management, New 
York: Society for Advancement of Management, 1951, 129-38. 

76. H. C. Warren, Dictionary of Psychology. Houghton Mifflin Co., 1934. 

77. J. B. Watson, Behaviorism. W. W. Norton & Company, Inc., 1925. 

78. T. N. Whitehead, Leadership in a Free Society. Oxford University Press, 1936. 

79. T. N. Whitehead, "Social motives in economic activities," Occup. Psychol^ 1938, 
12, 271-90. 

80. W. H. Whyte, Jr., The social engineer," Fortune, (January) 1952, 88 ff. 

81. W. H. Whyte, Jr., "Groupthink," Fortune, (January) 1952, 114ff. 

82. W. Williams, What's in the Worker's Mind. Charles Scribner's Sons, 1920. 

83. W. Williams, Mainsprings of Men. Charles Scribner's Sons, 1925. 

84. W. Williams, Full Up and Fed Up. Charles Scribner's Sons, 1925. 

85. R. S. Woodworth, Dynamic Psychology. Columbia University Press, 1918. 

86. J. B. Wolfe, "Effectiveness of token rewards for chimpanzees/' Comp. Psychol 
Monogr., 1936, 12, No. 60, 

The Impact of Social Psychology 
upon the Investigation of Moti- 
vation and Morale in Industry 

CONSIDERATION of human experience in the satisfaction 
of needs, wants and desires shows that "motivated behavior may move 
forward smoothly to its goal or it may encounter obstacles. The obstacles 
are sometimes readily surmounted, permitting satisfactory gratification 
of the desire. In other instances some degree of compromise or indirection 
provides a way out. In not a few cases, however, more or less persistent 
frustration occurs, accompanied by various only partially successful ad- 
justments" (84, p. 218). 

As is apparent from the above quotation from Kornhauser, something 
more than the classification of motives is needed to arrive at an under- 
standing of why people behave or fail to behave in a certain way. It is 
necessary, for example, to know something about the processes involved 
in setting goals. Consideration must be given to the influence of barriers 
to achievement whether these be actual obstacles existing in the real 
world or psychological states in the individual which in one way or an- 
other interfere with the release of internal tension and the satisfaction 
of basic needs or wants. In so far as the failure to achieve gratification 
of desires or wants may produce frustration, it is necessary to consider 
the mechanisms of frustration and associated effects in the way of ag- 
gression, fixation, etc. 

Those mentioned above and others to be discussed below represent, in 
fact, concepts derived from various fields of psychology which have been 
widely used in the consideration of motivation and morale in industry. 



Basic concepts and constructs developed from research in experimental 
psychology, clinical psychology, psychoanalytic psychology, animal psy- 
chology, and other fields have been effectively applied in the formulation 
of hypotheses and in the design of investigations undertaken to provide 
more complete understanding of what motivates workers to higher pro- 
duction, and of the determinants of individual satisfaction and morale in 

It is obviously impossible to present, within the limits of this volume, 
a systematic and detailed discussion of all the concepts derived from 
laboratory studies and experiments which have a bearing on motivation 
and morale. On the other hand, it seems quite necessary to discuss a num- 
ber of concepts which are referred to with some frequency in the discus- 
sion of the findings of experiments and of employee-attitude surveys de- 
scribed in later chapters. Furthermore, it seems appropriate, in discussing 
these, to adopt a frame of reference which is essentially that of social 
psychology, since the problems of motivation and morale in industry are 
closely bound to group situations. 


A survey of the development of social psychology will show that virtu- 
ally no field of psychology of/ as a matter of fact, of anthropology, soci- 
ology and, to some extent, of economics, has been overlooked by the psy- 
chologist concerned with the origins and explanation of social behavior 1 
(47). Simultaneously, social scientists in other areas have applied the 
theory, facts, and methods of psychology in dealing with problems which 
are described as being primarily anthropological, sociological, or economic 
in character. This trend is well illustrated in the early work of sociologists 
such as Durkeim (44), Ross (136), Thomas (150), Cooley (35, 36), and 
later in the work of other social scientists. 2 

Consideration of basic psychological concepts in the framework of 

1 Standard texts in the field of social psychology to which reference has been made 
in the preparation of this section include publications by McDougall (103), Allport 
(4), Bird (15), Dunlap (43), Young (161), Folsom (49), Murphy, Murphy and 
Newcomb (121), Klineberg (80), Britt (18), Bogardus (16), Newcomb and Hartley 
(124), Krech and Crutchfield (85), Sherif (144), Vaughn (151), Newcomb (123), 
Hartley and Hartley (68), and Asch (7). 

* Illustrative references include publications by G. H. Mead (114), Benedict (14), 
Malinowski (109), M. Mead (115, 116), Mayo (112, 113), theLynds (98,99, 100 , 
Dollard (39), Davis, Gardner and Gardner (37), Roethlisberger (130), Whyte 
(159), Gardner (58), Warner and his associates (153, 154), West (157), Leigh- 
ton (89), Greenwood (62), LaPiere (88), LaPiere and Farnsworth (87), Chappie 
and his associates (32, 33), Linton (96), Kardiner (77), Bakke (9), Chase (34), 
Dracker (42), Katona (78), and Miller and Form (118). 


social psychology (and of allied social sciences) is dictated by the fact 
that the problems of motivation and morale in industry are so predomi- 
nantly problems of social interaction of the type with which all social 
scientists are concerned. This does not mean that the study of the indi- 
vidual including individual differences is discarded when attention is 
focused upon motivation and morale in group situations. Orientation 
from the viewpoint of social psychology represents merely a recognition 
of the oft-observed fact that the individual may (and does) act differ- 
ently as a member of a group than when he is alone, and differently as 
a member of one group than as a member of some other group with which 
he is associated (67). 

In a factory, for example, the worker is simultaneously a member of a 
small work group or team, of a division or department, and of the or- 
ganization as a whole. In addition, under current conditions, he is very 
likely to be a member of a local union which may also have national and 
international associations. Within the factory, the worker may also belong 
to an employees' association, a bowling team, and one or more of many 
other groups. His attitudes, his morale, his goals and level of aspiration, 
his behavior at any moment may be predominantly influenced by any 
one of these associations. To examine the state of motivation and morale 
outside of the context of such group structures can readily result in an 
incorrect appraisal of the situation. Moreover, as Whyte has pointed out, 
while it may be necessary in research to focus attention upon the internal 
relations of a particular group, it is also necessary to be wary of general 
conclusions without examining also the "mutually dependent sets of rela- 
tions" among such groups (158). 

Motivation and Frustration 


As stated earlier, obstacles may be encountered as the individual pro- 
ceeds to the gratification of needs and wants (84) . Such obstacles to goal- 
oriented behavior may be in the form of internal barriers, Thus, deficient 
scholastic ability may block a highly motivated individual in the gratifi- 
cation of his "desire for prestige," to be achieved through the acquisition 
of an M.D. degree and the practice of medicine. 

Barriers to the gratification of needs, wants, desires, etc., can be ex- 
ternal in character. For example, an intellectually capable and highly 
motivated individual may be prevented from achieving the goal of medi- 
cal practice by financial handicaps, by resistance on the part of his family, 
or even because social norms of acceptability (as represented by prejudices 


applying to race or creed, etc.) prevent him from gaining admission to a 
medical school. 

The cases cited above refer to situations in which there is a single or 
predominant and highly positive goal which clearly patterns the sequence 
of behavior or goal response (40) to be carried through in satisfying under- 
lying motives. Obstacles to the complete satisfaction of needs and wants 
can arise out of the existence of conflicting goals. Both of these may be 
positive or desirable in nature, but movement in the direction of one will 
result in the satisfaction of one need while leaving another (or others) 
ungratified. An illustration of an individual "torn between conflicting 
desires" is found where a man must choose between accepting a "white- 
collar" job with low pay characterized by considerable prestige value, 
or a "blue-collar" r job with higher pay to which less prestige is attached. 
Frequently observed, also, is the conflict which arises in the presence of 
strongly opposed or even mildly contradictory goals and underlying de- 
sires. In industry, for example, the worker may view high pay as a desira- 
ble goal which, if achieved, would satisfy a basic drive toward a higher 
standard of living. He may simultaneously be faced with the need of 
reconciling this with the goal of restricting output as the only means for 
satisfying the need for status and recognition in the form of acceptance 
by a work group whose social, norms are directed toward the restriction 
of output 


There are many varieties of conflict situations (92). The blocking of 
goal-directed behavior and the accompanying failure to reduce tension 
produce a variety of consequences. Some of these are adaptive in char- 
acter and constructive in terms of outcomes (85). In the effort to "get 
around" or eliminate needs, the individual may arrive at some new way 
of solving a problem which both relieves the immediate tension and estab- 
lishes a pattern of behavior better suited to satisfy future needs. In this 
sense, personal handicaps and the vicissitudes imposed by physical, eco- 
nomic, and social barriers can reinforce motivation, strengthen indi- 
vidual character, and produce highly important results in the way of in- 
ventions, organization, or methods of dealing with problems, which have 
tremendous social consequences. 

On the other hand, extended and repeated thwarting of progress to- 
ward a goal and the persistent failure to resolve underlying tension can 
result in various kinds of maladaptive behavior. Among such possible 
consequences of frustration or "interference with the occurrence of an 
instigated goal-response at its proper time in the behavior sequence" (40, 
p. 7) are various forms of aggression. On the basis of relationships first 


postulated by Freud,* DoHard and others at the Institute of Human Re- 
lations, Yale University, have formulated a frustration-aggression hypoth- 
esis. This takes as the point of departure the assumption that "Aggres- 
sion is always a consequence of frustration. More specifically the proposi- 
tion is that the occurrence of aggressive behavior always presupposes the 
existence of frustration and, contrariwise, that the existence of frustration 
always leads to some form of aggression" (40, p. 1). 

It is to be noted that aggression, as used here, is not identical with the 
term "aggressiveness," but refers to an attack response associated with a 
"motive pattern directed toward inflicting injury" (123, p. 355). Feelings 
of anger, and hostile attitudes characterizing aggression and the attack 
conduct, may be and ordinarily are directed against the obstacle itself, i.e., 
the situation or person perceived as the source of frustration. For exam- 
ple, a child who is denied an ice cream cone strikes his mother with his 
fists; a worker who is denied an increase by his foreman "beats him up"; 
workers destroy machines which are perceived as causing unemployment, 

In the absence of an opportunity to attack the obstacle itself, the ag- 
gression may be displaced, in the sense of being diverted toward sub- 
stitute objects or persons, or even taking the form of acts of aggression 
towards oneself, i.e. self-injury. Thus, to use an example cited by New- 

* A summary of systematic contributions to the study of motivation and morale 
in industry would not be complete without at least a passing acknowledgment of 
the influence of psychoanalytic theory. The import of psychoanalytic theory has its 
origin in the point of view that a basis of social psychology is to be found in whatever 
contributes to an understanding of the determinants and dynamics of personality 
(5, 6, 149). Although the influence of Adler (1) and Jung (75, 76) is to be detected, 
the main line of influence is to be found in the work of Freud ( 52, 53, 54, 55, 56, 57) . 
While primarily concerned with the development and abnormalities of individual 
personality, Freud early took the position that "Only rarely and under certain ex- 
ceptional conditions is Individual Psychology in a position to disregard the relation 
of the individual to others. In the individual s mental life, someone else is invariably 
involved and so from the very first Individual Psychology is at the same time Social 
Psychology as well" (53). The most apparent contribution of psychoanalytic theory is 
in the form of terminology which has been widely adopted by psychologists concerned 
with the development of personality and with motivational theory. Thus, many of the 
concepts referred to in this section on the theoretical foundations of motivation and 
morale, such as frustration, aggression, fixation, ego, superego, etc., are borrowed from 
psychoanalysis. While seldom used specifically in their original sense by psychologists 
who have borrowed these terms, the debt to psychoanalysis is nevertheless clear-cut. 
Furthermore, as Landis (86) has suggested, psychoanalysis has served modern (includ- 
ing social) psychology (1) by forcing psychologists to inquire into the role of the 
infantile and early childhood fantasy on the mental life of the adult; (2) in indicating 
the dynamic role of the unconscious in individual behavior; (3) in calling attention 
to the significance of sex and sexuality and, in addition, by compelling psychologists 
to pursue the scientific study of personality as a step in setting up a theory and ex- 
planation of personality, and related theory of motivation, more valid and acceptable 
than those proposed by psychoanalysis (140). 


comb, "a frustrated boy who cannot find his baseball mitt because (as 
he believes) his brother has taken it, and who is prevented by fear of 
punishment from 'beating up' his brother, feels better after going to his 
brother's room and tearing up his bed'' (123, p. 355). Another widely 
used illustration with perhaps broader social significance, is derived from 
a study by Hovland and Sears (71 ) in which the annual number of lynch- 
ings of Negroes and property crimes with violence in 14 Southern states 
for the years 1882-1930 was used as measure of aggression. The annual 
per-acre value of cotton was taken as an index of economic conditions, 
with the further assumption that a low index represented a sign of greater 
interference with customary goal responses than did a high index (repre- 
senting good economic conditions). The correlation between this index 
and number of lynchings was found to be .67. In other words, the 
number of aggressive responses increased as the amount of interference 
with the attainment of an economic goal increased. According to Dol- 
lard et aL, this correlation indicates not only the functional relation be- 
tween frustration and aggression, but also that the aggression is displaced 
from the source of the frustration to the Negroes. "By no stretch of the 
imagination/' it is pointed out, "could it be assumed that the lynched 
Negroes were the source of the frustration represented by the low per 
acre value of cotton" (40, p. 44). 

As will appear in later sections, the aggression response to frustration 
has wide implications in dealing with specific problems of motivation 
and morale, and with the problem of industrial strife. In terms of broader 
social significance, the frustration-aggression principle is seen by the Yale 
group as underlying (at least by implication) the theories of the class- 
struggle and of the nature of the state in Marxian doctrine ( 1 10, p. 23 ) . 


In the frustration-aggression hypothesis referred to in preceding pages, 
as originally stated, aggression is viewed as the "primary and character- 
istic reaction to frustration" and will occur, in the form of one or another 
direct or substitute response, whenever there is interference with the 
motivated sequence of behavior. In this sense, as Newcomb points out, 
"frustration always induces motivation to some sort of aggression, 
and ... if no aggression in fact occurs, it has been inhibited" (123, p. 
353). Modifications and extensions of this theory, both by its original 
exponents 4 and by others, have led both to redefinitions of the term 
frustration and to the point of view that hostile attitudes and aggressive 
conduct are not the sole, or even the primary, sequelae of frustration. 
Thus, the position has been taken that "if . , . goal-directed behavior 

4 I.e. Bollard and his associates (as derived from Freud). 


is interfered with, or if the expected consequence of the activity is not 
forthcoming, the individual may experience deprivation, frustration, or 
a threat to his security. Of this," it is added, "the sense of depriva- 
tion is the least, and the sense of threat the most, disruptive to the indi- 
vidual" (68, p. 715). 

Not only are there varied definitions of the term frustration, but there 
is an increasing acceptance of the views that frustration leads to a variety 
of effects, and that aggression may or may not be the most common of 
these. In fact, members of the Yale group (119, 141) have revised their 
viewpoint "in such a way that nonaggressive reactions to frustration were 
reinstated in a position of proper importance" (102). Current literature 
provides a number of classifications in which nonaggressive reactions are 
assigned an appropriate role (101, 134, 142) and also reports on studies 
bearing upon the frequency with which these various types of reaction 
occur (41, 135). Illustrative of the findings of such studies are data from 
an investigation by McClelland and Apicella, involving 28 subjects who 
were subjected to "moderate" and "severe" frustration in the form of 
derogatory remarks in a laboratory situation (102). In this instance, the 
analyses of findings indicated that spontaneous comments made by sub- 
jects exposed to a blocking of desired goal responses could properly be clas- 
sified as representative of four "basic types" of frustration-instigated be- 
havior, viz., withdrawal, attack, limitation, and substitution. 


The work of Maier (107, 108) furnishes an illustration of a theory of 
frustration which includes a variety of frustration-instigated responses. 
The theory is posited upon a basic assumption that motivated behavior 
differs from frustration-instigated behavior. In terms of this distinction, 
motivated behavior is invariably goal-oriented. Conflicts between choice 
or preference are resolved, and needs satisfied, by the attainment of one 
or another goal. By contrast, according to Maier, "in frustration-instigated 
behavior there is no goal orientation." This theory leads necessarily to 
the distinction between two possible forms of satisfaction, viz., "relief 
from frustration, and the removal of a need through the attainment of 
a goal" (107, p. 99). 

According to this point of view, when a situation becomes frustrating, 
the behavior of the individual undergoes a complete change. An indi- 
vidual who was formerly unemotional and reasonable becomes emo- 
tional and unreasonable. Constructive, goal-directed behavior is replaced 
by stereotyped or compulsive, frequently pointless, and even destructive 
behavior. The major characteristics of such frustration-instigated re- 
sponses, as classified by Maier, are aggression, regression, fixation. To 


these is added resignation as a symptom frequently found in case-history 
studies (107). 

As in other classifications, aggression behavior is defined as a form of 
attack. This may be direct, in the form of physical violence. However, 
inhibitions developed by training generally lead to substitute responses, 
in the form of direct verbal attacks, sarcasm, gossip, etc. 

Regression is a "breakdown of constructive behavior and represents a 
return to childish behavior." According to this viewpoint, name calling 
by adults engaged in an argument is viewed as a common instance of re- 
gression. In industrial employees signs of regression include loss of emo- 
tional control, "horseplay," unreasoned fear, "following the leader" and 
similar reactions. "Men who pout, girls who cry easily, and workers who 
form childish cliques or gangs within the plant are expressing regressive 
behavior" (107, p. 64). 

Fixation is used to describe a "compulsion to continue a kind of ac- 
tivity which has no adaptive value." A common expression of fixation is 
an inability to accept change; a stubborn and unreasonable defense of old 
methods. According to Maier, "Industrial firms which are relatively free 
from frustrating situations and have high employee morale are made up 
of individuals who seek new ways rather than fear them" ( 107, p. 67) . 

Resignation is described as the process of "giving up." Such a reaction 
has been reported by Bakke (10), Eisenberg and Lazarsfeld (46), and 
others (31, 137, 162) in the study of changes in the personalities of 
unemployed individuals. "In industry," writes Maier, "the resigned indi- 
vidual is one who has lost hope of bettering his conditions" (107, p. 68). 


Responses of the type described above may be considered as pointless, 
at least from the viewpoint of correcting the situation which induced the 
frustration. However, they serve the purpose of providing relief from 
frustration and in this respect, according to Maier, such responses need 
not be viewed as being maladaptive in character. It is from this point of 
view that counseling programs and similar devices which provide em- 
ployees an opportunity to "blow off steam" and "express freely pent-up 
aggression tendencies" may be considered as having a beneficial effect 
(111, 131, 132, 17). Furthermore, in terms of Maier's assumptions, relief 
from frustration by release of the behavior which it instigates sets the 
stage for a return to constructive, goal-oriented or motivated behavior 
(108) . The basic problem, however, is still that of preventing, in so far as 
possible, the onset of frustration (93) . Newcomb, for example, has shown 
that frustration tends to fixate the attitudes prevailing at the time of 


frustration (122). Groups of frustrated individuals can become organized 
or united around a harmful pattern of aggression. Thus, the IWW (In- 
dustrial Workers of the World), active during the early part of the cen- 
tury, which had a history of intensive aggression, may be viewed as a labor 
organization which conforms to the criteria of a frustration-instigated 
social movement. "The effects of frustration/' writes Maier, "make possi- 
ble a fanatical type of social movement, in which the individuals are 
relatively homogenous and are dominated by hatred and destruction. 
They are persistent, irrational, and ready to follow a leader . . . Frustra- 
tion-instigated social movements thus constitute a social risk, since they 
are not oriented to a future goal" (107, p. 72) . 

A Positive Approach to the Avoidance of Frustration: Such effects, 
whether at the level of the individual plant or in a broader social sphere, 
point strongly to the need for a positive approach to the avoidance of 
frustration. This is to be found regardless of the theory of motivation 
and frustration which is accepted in providing conditions conducive 
to goal-oriented behavior. Adequate rewards positive incentives of all 
kinds are among such conditions. According to Maier the possible ef- 
fects of rewards are twofold. They may serve to reduce frustration by 
satisfying needs which, in the presence of continued deprivation, pro- 
duce frustration. Rewards may, in fact, translate behavior, in Maier's 
terms, "from the frustration process to the motivation process/' Secondly, 
where a state of motivation exists, rewards tend to maintain and enhance 
that state (108, p. 196). 

A major problem in research bearing upon motivation in industry is 
that of determining the extent to which various incentives and other con- 
ditions contribute toward the stimulation and continuation of behavior 
oriented toward a goal which both satisfies the needs of workers and the 
necessities of the industrial situation. The significance of experiments 
described in later chapters is derived, in part, from what they show con- 
cerning the role of potential rewards in the way of higher wages, praise, 
job security, etc., both in avoiding frustration and in maintaining and 
raising the level of aspiration in goal-oriented behavior. 

Sources of Frustration in Industry: Recent years have witnessed an in- 
creasing emphasis upon frustration as a key issue in dealing with human 
problems in industry, and with the social problems of industrial strife. 
Underlying this viewpoint is the position that "the job conditions under 
which people work are responsible, directly or indirectly, for the crea- 
tion of many tensions and the frustration of many needs and demands" 
(85, p. 539). These tensions, it is claimed, result in part from technologi- 
cal changes (117) which "frequently operate so as to block the expres- 
sion of many important needs of the worker on the job" (85, p. 539). 


Frustration and associated tension are also ascribed to the character of 
interpersonal relationships in modern industrial plants. Thus, according 
to Krech and Crutchfield, the indirectness of contact between top man- 
agement and the worker both contributes to the frustration of needs and 
to the worker's feeling that he is being "pushed around'' by manage- 
ment (85, p. 540) . The failure to provide an adequate income, with con- 
sequent blocking of needs and demands outside of the plant, is also 
viewed as a source of frustration off the job which "frequently intensifies 
the frustration of needs on the job ... Where the lack of money pre- 
vents the expression of needs off the job and the job situation does not 
permit their expression at work, then we have/' it is claimed, "a com- 
pounding effect all of it traceable to the job situation" (85, p. 540) . 

Still other possible sources of frustration in industry are included in 
a listing, by Eaton, of seven hypotheses concerning the origins of frustra- 
tion in industry to which research should be directed (45) : 

Hypothesis I: The worker is frustrated by the insignificance of 
his -work. 

Hypothesis II: The worker is frustrated by absentee ownership of 
the production of his work. 

Hypothesis III: The worker is frustrated by the [unfulfilled] ex- 
pectations of upward mobility which attend his 

Hypothesis IV: The worker is frustrated by his lack of a defined 
role, and by the many alternatives available in his 

Hypothesis V: The worker is frustrated by the changing tech- 
niques and conditions of his work. 

Hypothesis VI: The worker is frustrated by the isolation of his 
work within the community. 

Hypothesis VII: The worker is frustrated by the economic inse- 
curity of his work. 

The Role of the Union: One criticism directed against industry is that 
management does not voluntarily do all that is within its power to re- 
move such sources of frustration and thereby to provide enlarged op- 
portunities for the satisfaction of workers' needs. The strength of the 
labor union, according to this point of view, lies in the fact that the 
worker can look to labor organizations for support in eliminating ob- 
stacles to the gratification of needs which are not removed by man- 
agement. Increases in wages and improvement in working conditions 
achieved through the union are seen as accomplishments in this direc- 
tion. It is from this point of view that "aggression" behavior in the form 
of strikes for increased wages, better working conditions, shorter hours, 
etc., is viewed by Krech and Crutchfield as adaptive conduct, even 


though it may lead to industrial strife (85, p. 571). In a similar vein, 
Maier views continued opposition by management to goal-oriented labor 
unions as an instrument for increasing frustration and converting labor 
unions into militant, antisocial organizations (104, p. 74). 

It has been further suggested that management does not entirely have 
it within its power altogether to eliminate obstacles to the satisfaction of 
needs and wants in industry. So, for example, even with the best of in- 
tentions, management cannot completely eliminate repetitive work, dirty 
work, etc., or raise wages of all workers to the point where all their needs 
outside of the plant will be gratified. As a result, workers must seek "sub- 
stitute" satisfaction outside of industry. According to Krech and Crutch- 
field, the chief medium through which workers can obtain such sub- 
stitute satisfaction is the labor union. The latter, they say, "can better 
meet most of the worker's needs and demands than can other organiza- 
tions" in modern society (85, p. 548) . 

The issue raised above is discussed in some detail in later chapters. 
For the time being, it seems necessary only to note that this portrayal 
of the labor union is certainly far from acceptable to management and 
also to a substantial segment of the general public. Involved here, in 
large part, are differences in the way of perceiving a situation which are 
associated with group membership and which reflect the operation of 
socially developed norms. The nature and origin of such differences, and 
their effects on employee motivation and morale, cannot be understood 
without a consideration of social perception, social norms, and of related 

"Social Perception" and "Social Norms" 

According to Hastorf and Knutson, many of the basic problems of at- 
titude formation, persistence, and change can be understood only if it is 
recognized that "our individual and social purposes are important deter- 
miners of what is perceived/' 

". . . differences in the evaluation of a social situation are not differ- 
ences in conscious judgment and interpretation alone. Prior to these 
conscious judgments and interpretations, there very well may be more 
basic differences in the manner in which situations are perceived, de- 
pending on differences in past experience and purpose. Fundamental 
attitude changes are not likely to occur unless there are changes in the 
individual and social purposes which often play an unconscious selec- 
tive role in perception . . . 

"The attorneys for the N.A.M. and the C.I.O. in debating about the 
provisions of the Taft-Hartley Bill and the probable effect of its passage 
on the public welfare really have no parallel topics for debate. They 
neither perceive the same Taft-Hartley Bill nor the same public welfare. 


Since the disagreements are more fundamental than differences in con- 
scious judgments or interpretations, no amount of debate is likely to be 
effective. These judgments or interpretations may change, but they will 
still remain judgments or interpretations of differently perceived situa- 
tions, unless one or the other of the attorneys first undergoes a complete 
breakdown and reorientation of his identifications and related striv- 
ings . . ."(68a,p. 234). 

In the above quotation is found an illustration of the current treat- 
ment of perception by social psychologists which is of significance in the 
consideration of motivation and morale in industry. Perception, as tradi- 
tionally defined, refers to the "awareness of external objects, qualities, 
or relations, which ensues directly upon sensory processes" (155, p. 196) . 
In the simplest terms, perception is the process through which, for ex- 
ample, a pattern of lines and dots on a piece of paper is observed as a 
face, a combination of auditory stimuli as a melody or tune, etc. Signifi- 
cant research conducted by the Gestdt psychologists, as represented par- 
ticularly in the pioneering work of Kohler (82, 83), Koffka (81) and 
Wertheimer (156), has shown that ''the qualitatively distinct character 
of perception of form, melody, rhythm, meaning, is not derived from 
the distinct properties of the parts in isolation, but that (on the contrary) 
the parts derive their quality from their functional membership in the 
whole" (144, p. 157). Furthermore, it has long been recognized that the 
attainment of meaning in perception is achieved, in part, through an 
application of the past experience of the individual to the interpretation 
of objective stimuli in the external world. For example, according to 

"If a person in a dark room sees two illuminated toy balloons, one 
of which expands as the other gets smaller, he will interpret this not as 
a change in the size of the balloons, but as a change in their distance 
from him. The one that expands will seem to be coming toward him; the 
one that contracts will seem to be rushing away from him. What the 
individual experiences, then, is movement of the balloons. 

"Why does he experience movement instead of a change in size? 
The apparent answer is that from past experiences the assumption is 
made that two things (two toy balloons) which are similar are likely 
to be identical. And from our past experience we know that if two things 
are identical and one is larger than the other, it should be closer to us. 
And so when one balloon is made larger, we see it coming closer while* 
the one made smaller recedes" (25, p. 51). 

The process of perception was at first treated by psychologists as one 
of interest only from the viewpoint of the individual functioning in 
isolation. More recently, research by Sherif (145, 146) on group influences 
on the perception of the autokinetic effect (i.e., of the apparent move- 
ment of stationary objects); by Bartlett in the field of "remembering" 


(11); by Hallowell on the personality of primitive man (64, 65); by 
Brunei and Goodman on the perception of size of coins by "poor" and 
"rich" children (21); and studies of the perceptual process in varied con- 
texts by Murphy and others (90, 91, 143); by Postman and his associates 
(2, 126, 127), etc., have turned attention to the problem of the social 
determination of perception and the related problem of the perception 
of the social (106). Such research has led to the conclusion, by Hartley 
and Hartley, that "the process of perception itself is susceptible to social 
influence and in turn largely determines social behavior" (68, p. 226) . 


According to Goodenough and Anderson (60), Sherif (147), and 
others, the field of perception has become a common meeting ground 
for social scientists and "perceptual reactions have become the prototype 
of all psychological reactions for social psychologists in singling out the 
effects of socio-cultural influences on the one hand and biological and 
strictly personal factors on the other" (147, p. 11). Of particular signifi- 
cance in this approach is the emphasis placed upon the role of group- 
determined frames of reference or of social norms upon the attitudes 
and performance of the individual. Underlying this approach is the view 
that "all perception is to a lesser or greater degree social perception" 
since, according to Bruner and Postman, "the set which the individual 
brings to a perceptual situation is a function of his prevailing motives, 
needs, attitudes, and personality structure all of which, in turn, are 
products of the interaction between the organism and his social environ- 
ment" (20, p. 71). 

It is not within the province of this volume to examine the detailed 
mechanisms by which social perceptions arise, the processes through 
which perception influences social behavior, and the way in which "social 
norms" are developed in the operation of perception and of other mental 
processes. The reader who is interested in a detailed treatment of such 
psychological processes is referred to Kelson's treatment of quantitative 
methods for the prediction of frames of reference (69) and to the de- 
tailed consideration of social norms in texts by Sherif (144), Rohrer and 
Sherif (133), Krech and Crutchfield (85), Cantril (25), Hartley and 
Hartley (68), and others. However, a few illustrations of the operation 
of "social perceptions" and "social norms" may help underline the 
significance of these concepts, and facilitate the interpretation of ex- 
periments and attitude surveys bearing upon motivation and morale 
described in later chapters of this volume. 

The import of culturally determined norms in evaluating man's psycho- 
logical responses to the physical objects of his external environment (66) 


can be illustrated by reference to an observation of a primitive culture 
by Linton. He observed that in one of the small tribes of the southeast 
coast of Madagascar, iron cooking pots were owned by most families. 
These had been introduced about twenty years before his visit by men 
who had returned to the tribe after working as contract laborers. To do 
work of this type conferred prestige on a man and the iron pots, purchased 
from one's earnings, became a durable symbol of that prestige. To have 
a long row of iron pots on display "had become a symbol of the social 
importance of the household" (97, p. 477). 


It is apparent th^t the meaning attached to the pots extends beyond 
their usefulness as cooking instruments. This illustration shows clearly 
that an object is seen not merely as possessing form, color, and utility, 
but is viewed or perceived by members of a particular culture or of a 
specific group as having certain socially significant connotations. Differ- 
ences in norms of evaluation among smaller groups within the same 
culture are apparent, for example, in the contrasting value attached to 
old, rebuilt, and sometimes highly decorated "hot rods" by teen-agers 
and adults respectively within the American scene. In industry, as has 
been pointed out by Roethlisberger (130) and Gardner (58), the physi- 
cal location of an office, office furniture, floor carpeting, ash trays, etc., 
are not seen merely as physical appurtenances of the work situation, but 
as symbols of status and prestige. The Rank-Has-Its-Privileges Execu- 
tive Evaluation chart, duplicated on page 448, which has been widely 
passed around in management circles in the United States, strikingly 
reveals the significance actually attached to such physical objects as status 
symbols in the social microcosm represented by the modern industrial 

As has been suggested earlier, one product of social interaction is 
the development of norms which reflect a group-determined frame of 
reference in viewing a situation. Such social norms appear along with 
the formation of a group, as is illustrated in the new and higher norms of 
production established along with the growth of the "informal social 
organization" in the Relay Assembly Test Room, in studies conducted in 
the Hawthorne Works of the Western Electric Company, discussed in 
Chapters 10 and 11. Once established, such norms of conduct by the 
group can persist for a long time, as illustrated in the restriction of pro- 
duction by the Bank Wiring Group observed in the same series of in- 
vestigations, and also in the numerous instances of restriction of out- 
put by group members referred to elsewhere in this volume. An important 
aspect of this situation is the fact, established in the laboratory study by 


Sherif, that "group norms enter to shape the reactions of the individual 
member even when he is no longer actually in his group'' (144, p. 177). 
In industry, it has been frequently observed that an individual carries 
such group-determined norms with him as he moves from one situation 
to another. This is well illustrated in a report, by Gardner, of the in- 
fluence of a group-established norm bearing upon the rate of learning 
by a worker as he moves into a new job. 

"One old-timer had been brought in from another totally different 
job and put on sand-blasting. He had had experience on sand-blasting 
years before and picked up the work very rapidly. He watched his daily 
reports very carefully, however, and showed only a moderate increase 
each day even though he was well below what was considered a fair day's 
work, and was not working very hard. When asked about the work, he 
said that he could have reached the average output within three days but 
was careful not to progress too fast. He explained that the proper learn- 
ing curve for a new man on the job was one which showed fairly rapid 
improvement for about a, week and then a gradual slowing of progress so 
that it would take several weeks to reach average output. He was making 
his output conform to this proper pattern, since if he improved too fast, 
the foreman would know that the group were taking it fairly easy and 
would try to make them increase their output" (58, p. 158). 

Illustrations cited above reveal the intimate relations between "social 
perceptions" and "social norms" on the one hand, and motivation on the 
other. Thus, according to Newcomb, "Motivation is the concept in terms 
of which we describe the over-all consistency of overt-perceptual be- 
havior of each man. This consistency could not occur without selectivity 
in taking note of the environment. The 'outer' part of motivation is thus 
dependent upon perception" (123, p. 89) . 

In this sense, the strength of the "drive" exerted toward acquiring a 
particular make of automobile will be influenced by the perception of the 
vehicle as "large" or "small," and by the wider social frame of reference 
through which the status or prestige value of the particular make of car 
is evaluated. Similarly, the strength of the desire for higher wages (and 
the amount wanted) may be a function of the extent to which wages 
are viewed or perceived as placing workers in a group "in line" or "out 
of line" with area levels and with the earnings of other individuals. 5 
Furthermore, as Newcomb also points out, the relationship between 
social perception and motivation is circular in character, in that "selec- 
tivity in perception is determined by the existing state of motivation." 
In this sense, the individual may perceive a particular make of automobile 
as being particularly attractive or as having certain superior character- 
istics because he has already become highly motivated in terms of pur- 

See Chapter 15. 


chasing this make of car. Similarly, a worker already poorly motivated 
from the viewpoint of production may perceive many elements of the 
work situation in an entirely different context from that of the highly 
motivated worker. 

Ego, Status, and Role 

A problem of both theoretical and practipal importance in the con- 
sideration of social norms in relation to motivation centers around the 
question of the influence of such norms upon the way in which the indi- 
vidual regards himself and his needs. Stated in different terms, this ques- 
tion bears upon the nature and the origin of the ego, defined by Warren 
as "the individual's conception of himself (155, p. 89) and by Cantril 
as "a concept to describe what each individual subjectively regards as 
me" (24, p. 44) . Either of these appears acceptable as a "working" defini- 
tion for purposes of this text although, as has been made clear by All- 
port (3), there is far from uniform agreement among psychologists in 
dealing with this concept. 6 

With respect to the genesis of the ego, theories range from McDougall's 
conceptualization of self-regard as instinctively determined (104) to the 
view of the "ego as the subjective organization of culture." Somewhere 
midway between an extreme- emphasis upon "internal" and "external" 
factors, respectively, is the position taken by Cantril that "the ego of 
an individual is essentially composed of the many social and personal 
values he has accepted" T (22, p. 197). However, according to the latter, 
"since the ego for most individuals is so largely composed of social 
values, almost every person to some extent evaluates himself in terms of 
the norms of his particular society" (24, p. 42). At the same time, also 
according to Cantril, "a person's ego and, consequently, the way in which 
he regards himself, are by no means always entirely bound by the sur- 
rounding culture." Thus, the individual may cherish values shared by the 
family circle, a few professional colleagues, etc., more than the more 
common values of society conducive to social recognition. And some- 
times he may cherish, as most important of all, those values which he has 
worked out for himself and thinks of as his own (24, p. 44) . 

Although differences of opinion exist concerning the significance of 
social and personal factors in the genesis and development of the ego, 
there is common agreement that rewards or incentives can have but 
limited value as motivating forces except in so far as they cater to the 
satisfaction of ego4nvolved needs, and contribute to the maintenance 

Although derived from psychoanalysis, the term "ego" is not generally used by 
psychologists in the strictly psychoanalytic sense. 
T Not so italicized in the original. 


and enhancement of the ego. A practical problem in industry centers 
around the question of how to bring about this enhancement of self- 
regard in such a way that both the needs of the individual and the neces- 
sities of the industrial organization will be simultaneously served. Stated 
in other terms, the problem is to facilitate the development of a "con- 
stellation of ego-attitudes" (148) which will lead the individual to per- 
ceive greater production as his goal; to look upon the foreman as his 
foreman; to view the company as his company. Thus, according to Sherif 
and Cantril, the outstanding accomplishment of the Lincoln Electric 
Company incentive management plan (95, 152) may be ascribed to the 
fact that Lincoln "employees regard themselves more as 'business- 
men' . . . identify themselves with the company and its welfare, not 
with other working people in the larger world outside" ( 148, p. 373) . 

A crucial element in this situation is the character of the social norms 
existing in the miniature society represented by the Lincoln plant. It is 
the operation of social norms which leads the individual to consider his 
status to be that of "businessman" rather than a "hired hand" and brings 
about behavior sometimes identified as role behavior appropriate to 
his evaluation of himself in terms of the norms of this particular society. 
This illustration brings into relief a number of concepts, viz., position, 
status and role, which have been widely referred to in the literature of 
social psychology in dealing with the development of the "ego," and in 
grappling with problems of motivation and morale associated with social 
interaction in industry. 


Such problems, as they occur in industry, are discussed in later chap- 
ters of this text. Of broad interest, as a background in dealing with such 
problems, are investigations bearing upon the relation between status 
in terms of "class" identification and the attitudes, goals and behavior 
of "class-conscious" individuals (97a, 125a). Status, as used in this con- 
nection, refers specifically to the individual's rank or position in society 
as he sees it. 8 Studies of "class identification" or "class consciousness" by 
Kornhauser (84), Cantril (23), Hyman (72), Centers (27, 28, 29, 30), 
and others (125) are generally interpreted as supporting the conclusion 
that "the American people do identify themselves with a social class" 

8 There is a good deal of disagreement in the usage of the term "status" and of other 
concepts mentioned above. Thus, Sherif and Cantril (148), use the terms "role" and 
status" interchangeably, while Linton (96), makes specific distinctions between 
these. Similar contradictions apply to the use of the concept "position." For detailed 
consideration of the meanings attached to these terms, the reader is referred to 
Guthrie (63), Hartley (67), Sargent (138), and Hartley and Hartley (68). 


and, furthermore, that "this identification is highly correlated with the 
particular roles they play in a highly industrialized society" (148, p. 194). 
Thus, for example, a study by Cantril (23) showed that 99 per cent of a 
nation-wide sample of the American adult population gave a definite 
answer each member placing himself in a specific "social class" in 
replying to the question "To what class in this country do you feel you 
belong middle class, the upper, or lower?" Furthermore, in this case, 
approximately 90 per cent identified themselves with the "middle class/' 
In a later study, by Centers (26), approximately 1,100 individuals, 
representing a cross section of the adult white male population, were 
asked: "I/ you were asked to use one of these four names for your social 
class, which one would you say you belonged in: middle class, lower class, 
working class, or upper class?" Replies were distributed as shown in Table 
7. Significant in this distribution is the shift from self-classification as 
members of the "middle class" to that of the "working class" when the 
latter category is added to the list of classes. This shift toward identifica- 
tion with the "working class" and also "laboring class" has again been 
observed in later studies by Centers. 

TABLE 7. Percentages of the Population Affiliating with Each Social Class 

(after Centers) 

Per cent 

Upper class 3 

Middle class 43 

Working class 51 

Lower class 1 

Don't know 1 

Don't believe in classes 1 

The Influence of Socio-economic Position: A number of the investiga- 
tors referred to above have dealt with the criteria employed by indi- 
viduals in identifying themselves with one or another of the "social 
classes." Findings indicate clearly that self-placement within one or an- 
other social "class" is related to socio-economic position, particularly 
as reflected in the occupations of adults and, in the case of children, in 
the occupations of their parents (124a, p. 166 ff.). Thus, analysis of data 
referring to the sample of approximately 1,100 adults interviewed in the 
study by Centers (26, 29) showed that approximately 13 per cent of 
owners or managers of large businesses classified themselves as "upper 
class." Only 3 per cent of owners and managers of small businesses placed 
themselves in this category. Furthermore, no unskilled worker classified 
himself as "upper class." Similarly, approximately 75 per cent of owners 
or managers of large businesses classified themselves as "middle class" as 


contrasted with approximately 30 per cent of skilled manual workers who 
identified themselves with this social "class/' 

The individual's conception of his position with respect to "social 
class" is of importance, from the viewpoint of motivation and morale 
in industry, because his attitudes are related to his status as he sees it 
Furthermore, attitudes which are related to role or status are ego-involved 
(148). Examples of such relation between "class" status and attitudes 
toward wages, job security, unions, and other specific aspects of the in- 
dustrial situation are reported in later chapters. Moreover, Centers has 
shown that there are differences among "classes" with respect to attitudes 
toward broad political and social issues, such as government ownership 
of industry, America as a land of opportunity, the role of government in 
providing security, attitudes toward labor, etc. (26, 27). According to 
Centers, "an examination of these differences can leave little doubt that 
people's politico-economic orientations are closely associated with their 
statuses and roles in the economic order 9 ' (26, p. 482) . 

The above discussion has been concerned with the broad problem of 
status in relation to social "classes." The question of status is of equal 
and perhaps of greater importance in relation to smaller social groups. 
In the miniature society of the industrial plant, as well as in society at 
large, there are variations in status and associated behavior roles. In fact, 
one of the current problems in industry arises from difficulties experi- 
enced by employees in subjectively accepting a change in status when 
there has been an actual modification of position in the industrial hier- 
archy. This appears particularly in the case of first line supervision (50) . 

"Subjective" and "Objective" Status: In the preceding section status 
has been considered from the view of the position or rank of the indi- 
vidual in a society as he sees it. This can be referred to as the subjective 
orientation of status, as implied in Cantril's definition of status as "the 
relationship between a person's interiorized values and the norms of his 
society" (24, p. 42). Status may also be looked upon as the ranking or 
position assigned to an individual by others in the group with which he 
is associated. From this point of view, status might be considered as 
synonymous with "social recognition." It is partly in this sense that, as 
reported by Whyte (160), workers at the salad station in a restaurant 
are accorded higher status by fellow employees than those engaged in the 
preparation of vegetables or fish. 

Difficulties can arise out of the dichotomous aspects of status, be- 
cause of discrepancies between the individual's subjective evaluation of 
his status and the objective evaluation of his status by his associates 
or superiors within the framework of the social structure. Thus, for ex- 
ample, the foreman may look upon himself as occupying a key position 


in an industrial plan whereas he may, in fact, occupy a position of little 
responsibility or authority because of a management outlook which con- 
siders the foreman to be far from an indispensable cog in either its 
operating or industrial relations program. 


This problem is discussed in Chapter 23. For the purposes of this sec- 
tion, it seems desirable to turn to another and important aspect of status 
relationships specifically, the function of status in specifying the role 
behavior of individuals at various levels in the hierarchy of a social struc- 
ture. As Hartley and Hartley have pointed out, more or less specific forms 
of behavior are prescribed for the contacts between persons at different 
levels of a hierarchy of status. "Status systems customarily define the 
pattern of relationships which govern interaction among group members" 
(68, p. 572). Behavior roles may be graded with regard to the extent to 
which the role behavior is determined by consideration of status. Thus, 
the accepted custom of deferring to the "boss" in tone of voice, in order 
of entering a room, in a "thousand other more subtle details of behavior" 
is an obvious example of "ordering behavior to status demands" (68, p. 
573). How status affects relationship and defines roles in industry ap- 
pears, according to Gardner, in an entire series of man-boss relationships 
in which 

"each person is intensely concerned with how his boss judges him and 
at the same time is busy judging his subordinates. Each is constantly 
looking at his subordinates, trying to determine how well they are doing 
their jobs, and how they might do better work, and each is constantly 
being irritated and disturbed when they fall short of what he thinks they 
should be doing. At the same time his concept of the job is constantly 
being mixed up, with what his boss will think and what he expects, 
until 'doing a job' often becomes a matter of 'doing what the boss thinks 
is good/ Often this concern is not merely with what the boss expects in 
terms of the work itself, but also with what he thinks is 'proper' be- 
havior. As a result each level is constantly judging his subordinates not 
merely in terms of the work accomplished but in terms of 'what would 
my boss think if he saw them?' " (58, p. 9). 

The above illustration suggests that the choice of role behavior which 
is "situationally appropriate" presents difficulties even when the worker 
is concerned only with his status in relation to that of his immediate 
supervisor. This situation is further complicated by the fact that the indi- 
vidual worker is simultaneously a member of a number of groups which, 
because of varying group expectations and demands (138), impose dif- 
ferent and at times contradictory roles upon him (67). According to 
Linton, "in those cases in which . . . statuses whose roles are funda- 


mentally incompatible converge upon the same individual, we have the 
material of high tragedy" (96, p. 80). Linton goes on to point out that 
such conflicts rarely arise in primitive societies, but become fairly fre- 
quent under conditions existing in modern society. In an era of rapid 
technological and social change, the individual "finds himself frequently 
confronted by situations in which he is uncertain both of his statuses 
and roles and of those of others" (96, p. 81). The decisions which he 
makes in terms of the behavior to be carried through may prove to be 
"out of line" with reciprocal expectations, with resulting disappointments 
and frustrations. 

In the industrial plant, the worker faces the necessity imposed by man- 
agement of being a high producer. Simultaneously, he may be presented 
with expectations on the part of his work team in the way of behavior 
designed to restrict production. He can obtain ego-satisfaction by "enact- 
ing the role" of "a loyal" employee, but may simultaneously be faced 
with conflicting demands for contradictory behavior by his union. Hav- 
ing improved his position (in the formal structure of the plant organiza- 
tion) by promotion to the job of foreman, he is faced with the risk of 
"losing status" among his former associates if his behavior is "situationally 
appropriate" to his new role. 

Such conflicting demands account for many of the difficulties in in- 
creasing motivation and raising morale in industry. Nevertheless, as will 
appear in later chapters, such devices as good supervision, employee par- 
ticipation in decision making, etc., serve to promote ego-involvement 
with a plant objective; raise levels of aspiration; 9 induce role behavior 
which is desirable from the viewpoint of achieving the objectives of 
industry. It is from this point of view that studies described in later sec- 
tions of this volume will be of particular interest to management 


1. A. Adler, The Practice and Theory of Individual Psychology. Harcourt, Brace & 
Co., 1924. 

2. G. W. Allport and L. Postman, The Psychology of Rumor. Henry Holt & Co., 

3. G. W. Allport, "The ego in contemporary psychology/' Psychol. Rev., 1943, 50, 

The fact that ego-involvement raises the level of aspiration has been demonstrated 
in a number of studies. The reader interested in contributions from this area from 
systematic psychology is referred to publications by Hoppe (70), Dembo (38), Lewin 
(92), Frank (51), Gould (61), Preston and Bayton (129), Irwin (73, 74), Festinger 
(48), Sears (139), McGehee (105), Bayton (13), Lewin et d. (94), Preston, Spiers, 
and Trasoff (128), and Bayton and White (12). 


4. F. H. Allport, Social Psychology. Houghton Mifflin Co., 1924. 

5. G. W. Allport, Personality: A Psychological Interpretation. Henry Holt & Co., 

6. A. Angyal, "A theoretical model for personality," /. Personal., 1951, 20, 130-42. 

7. S. Asch, Social Psychology. Prentice-Hall, Inc., 1952. 

8. Associated Research, Institute for: see 68a. 

9. E. W. Bakke, The Bonds of Organization: An Appraisal of Corporate Human 
Relations. Harper & Brothers, 1950. 

10. E, W. Bakke, Citizens Without Work. Yale University Press, 1940. 

11. F. C. Bartlett, Remembering: A Study in Experimental and Social Psychology. 
Cambridge University Press, 1952. 

12. J. A. Bayton and E. C. Whyte, "Personality dynamics during success-failure se- 
quence," /. Abn. Soc. PsychoL, 1950, 45, 583-91. 

13. J. A. Bayton, "Interrelations between levels of aspiration, performance, and 
estimates of past- performance," /. Exper. Psychol. t 1943, 33, 1-21. 

14. R. Benedict, Patterns of Culture. Houghton Mifflin Co., 1934. 

15. C. Bird, Social Psychology. D. Appleton-Century Co., 1940. 

16. E. S. Bogardus, Fundamentals of Social Psychology. D. Appleton-Century Co., 
1942 (3rd Edition). 

17. E. M. Bowler and F. T. Dawson, Counseling Employees. Prentice-Hall, Inc., 

18. S. H. Britt, Social Psychology of Modern Life. Farrar & Rinehart, Inc., 1941. 

19. J. F. Brown, Psychology and the Social Order. McGraw-Hill Book Co., 1936. 

20. }. S. Bruner and L. Postman, "An approach to social perception," in Current 
Trends in Social Psychology. University of Pittsburgh Press, 1948, 71-118. 

21. J. S. Bruner, and C. G. Goodman, "Value and need an organizing factor in per- 
ception," /. Abn. Soc. Psychol., 1947, 42, 33-44. 

22. H. Cantril, The Invasion from Mars, Princeton University Press, 1940. 

23. H. Cantril, "Identification with social and economic class," /. Abn. Soc. Psychol., 
1943, 38, 74-80. 

24. H. Cantril, The Psychology of Social Movements. John Wiley & Sons, Inc., 1941. 

25. H. Cantril, The "Why" of Man's Experience. The Macmillan Co., 1950. 

26. R. Centers, "American class structure: a psychological analysis" in T. M. New- 
comb and R. E. Hartley (Eds.), Readings in Social Psychology. Henry Holt & 
Co., 1947, 481-93. 

27. R. Centers, "Children of the New Deal: social stratification and adolescent 
attitudes," Int. J. Opin. Att. Res., 1950, 4, 315-35. 

28. R. Centers, "Nominal variations and class identification: the working and the 
laboring classes," /. Abn. Soc. Psychol., 1950, 45, 195-215. 

29. R. Centers, The Psychology of Social Classes. Princeton University Press, 1949. 

30. R. Centers, "Social class identifications of American youth," /. Personal, 1950, 
18, 290-302. 

31. S. N. F. Chant and R. C. Myers, "An approach to the measurement of mental 
health," Amer. J. Orthopsychiatry, 1936, 6, 134-40. 

32. E. D. Chappie (with C. M. Arensberg), "Measuring Human Relations," Gen. 
Psycho/. Monog., No. 22, 1940. 

33. E. D. Chappie and C. S. Coon, Principles of Anthropology, Henry Holt & Co., 

34. S. Chase, The Proper Study of Mankind. Harper and Brothers, 1948. 


35. C. H. Cooley, Human Nature and the Social Order. Appleton-Century Co., 1925. 

36. C. H. Cooley, Social Organization. Charles Scribner's Sons, 1909. 

37. A. Davis, B. B. Gardner and M. R. Gardner, Deep South. A Social Anthropologi- 
cal Study of Caste and Class. University of Chicago Press, 1941. 

38. T. Dembo, "Der Aerger als dynamisches Problem," Psychol. Forsch., 1931, 15, 

39. J. Bollard, Caste and Class in a Southern Town. Yale University Press, 1937. 

40. J. Dollard, L. W. Doob, N. E. Miller, O. H. Mowrer and R. R. Sears, Frustra- 
tion and Aggression. Yale University Press, 1939. 

41. L. W. Doob, and R. R. Sears, "Factors determining substitute behavior and the 
overt expression of aggression/' /. Abn. Soc. Psychol., 1939, 4, 293-313. 

42. P. F. Drucker, The New Society. Harper & Brothers, 1949. 

43. K. Dunlap, Social Psychology. The Williams and Wilkins Co., 1927. 

44. E. Durkheim, The Rules of Sociological Method. University of Chicago Press, 
1938 (8th Edition). 

45. W. H. Eaton, "Hypotheses related to worker frustration," /. Soc. Psycho/., 
1952, 35, 59-68. 

46. P. Eisenberg and P. F. Lazarsfeld, "The psychological effects of unemployment," 
Psychol. Bull, 1938, 35, 358-90. 

47. E. Faris, 'The beginnings of social psychology," Amer. J. Sociol., 1944-45, 

50, 422-28. 

48. L. Festinger, "Wish, expectation and group standards as factors influencing 
level of aspiration/' /. Abn. Soc. Psychol., 1942, 37, 184-210. 

49. }. K. Folsom, Social Psychology: Its Origins, Development and European Back- 
ground. McGraw-Hill Book Co., 1932. 

50. Foreman Thinking Today. Princeton: Opinion Research Corporation, 1946. 

51. }. Frank, "Individual differences in certain aspects of the level of aspiration," 
Amer. J. Psychol, 1935, 47, 119-28. 

52. S. Freud, "Mourning and melancholia" reprinted in Collected Papers: Hogarth 
Press, Ltd., 1934. 

53. S. Freud, Group Psychology and the Analysis of the Individual. Boni & Liveright, 


54. S. Freud, Civilization and Its Discontents. Hogarth Press, Ltd., 1930. 

55. S. Freud, Psychopathology of Everyday Life. The Macmillan Co., 1914. 

56. S. Freud, New Introductory Lectures to Psychoanalysis. W. W. Norton & Com- 
pany, Inc., 1933. 

57. S. Freud, A General Introduction to Psychoanalysis. Boni & Liveright, 1920. 

58. B. B. Gardner, Human Relations in Industry. R. D. Irwin, 1946. 

59. B. B. Gardner, "The factory as a social system" in W. F. Whyte (Ed.), Industry 
and Society. McGraw-Hill Book Co., 1946, 4-20. 

60. F. L. Goodenough and J. L. Anderson, "Psychology and anthropology: some 
problems of joint import for the two fields," Southwestern Anthrop., 1947, 3, 

61. R. Gould, "An Experimental Analysis of Level of Aspiration/' Genet. Psychol. 
Monog., 21, No. 1, 1939. 

62. E. Greenwood, Experimental Sociology. A Study in Method. King's Crown 
Press, 1945. 

63. D. Guthrie, The Psychology of Human Conflict. Harper & Brothers, 1948. 

64. A. I. Hallowell, "Personality structure and the evaluation of man," Amer. An- 
thropol^ 1950, 52, 159-73. 


65. A. I. Hallowell, "Cultural factors in the stnicturalization of perception" in J. H. 
Rohrer and M. Sherif (Eds.), Social Psychology at the Crossroads. Harper & 
Brothers, 1951, 164-95. 

66. A. I. Hallowell, "Psychological needs for ethnological workers" in D. G. Haring, 
Personal Character and Cultural Milieu. Syracuse University Press, 1949. 

67. E. Hartley, "Psychological problems of multiple group membership" in J. H. 
Rohrer and M. Sherif (Eds.), Social Psychology at the Crossroads. Harper & 
Brothers, 1951, 371-87. 

68. F. L. Hartley and R. E. Hartley, Fundamentals of Social Psychology. Alfred A. 
Knopf, Inc., 1952. 

68a. A. H. Hastorf and A. L. Knutson, "The nature of attitude and opinion" in F. P. 
Kilpatrick, Human Behavior from the Transactional Point of View. Hanover, 
N.H.: Institute for Associated Research, 1952, 233-37. (See also A. H. Hastorf 
and A. L. Knutson, "Motivation, perception and attitude change/' Psychol. Rev., 
1949, 56, 88-94.) 

69. H. Helson, "Adaptation-level as a basis for a quantitative theory of frames of 
reference," Amer. J. Psychol, 1942, 55, 297-313. 

70. F. Hoppe, "Erfolg und Miserfolg," Psychol Forsch., 1930, 14, 1-62. 

71. C. I. Hovland and R. R. Sears, "Minor Studies of Aggression: VI Correlation 
of Lynchings with Economic Indices," /. Psychol, 1940, 9, 301-10. 

72. H. H. Hyman, "The Psychology of Status," Arch. Psychol, No. 269, 1942. 

73. F. W. Irwin, 'The realism of expectation," Psychol Rev., 1944, 51, 120-26. 

74. F. W. Irwin, "Motivation" in H. Helson, Theoretical Foundations of Psychology. 
D. Van Nostrand Co., 1951, 200-253. 

75. C. C. Jung, Psychological Types, Harcourt, Brace & Co., 1923. 

76. C. C. Jung, Modern Man in Search of a Soul Harcourt, Brace & Co., 1933. 

77. A. Kardiner, and others, The Psychological Frontiers of Society. Columbia Uni- 
versity Press, 1945. 

78. G. Katona, Psychological Analysis of Economic Behavior. McGraw-Hill Book 
Co., 1951. 

79. D. Katz and R. C. Schank, Social Psychology. John Wiley & Sons, Inc., 1938. 

80. O. Klineberg, Social Psychology. Henry Holt & Co., 1940. 

81. K. Koffka, Principles of Gestalt Psychology. Harcourt, Brace & Co., 1935. 

82. W. Kohler, Gestalt Psychology. Liveright, 1929. 

83. W. Kohler, Dynamics in Psychology. Liveright, 1940. 

84. A. W. Kornhauser, "Analysis of 'class' structure of contemporary American 
society" in G. W. Hartmann and T. Newcomb (Eds.), Industrial Conflict, The 
Cordon Co., 1939, 199-264. 

85. D. Krech and R. S. Crutchfield, Theory and Problems of Social Psychology. 
McGraw-Hill Book Co., 1948. 

86. C. Landis, "Psychoanalysis and scientific method" in Symposium on Recent 
Advances in Psychology. Philadelphia: Proceedings of the American Philosoph- 
ical Society, 1941, 524-25. 

87. R. T. LaPiere and P. R. Farasworth, Social Psychology. McGraw-Hill Book Co., 
1942 (2nd edition). 

88. R. T. LaPiere, Sociology. McGraw-Hill Book Co., 1946. 

89. A. H. Leighton, The Governing of Men. Princeton University Press, 1945. 

90. J. Levine, and G. Murphy, "The learning and forgetting of controversial ma- 
terial," /. Afcn, Soc. Psychol, 1943, 37, 547 S. 

91. R~ Lcvinc, I. Chein and G. Murphy, 'The relation of the intensity of a need 


to the amount of perceptual distortion, a preliminary report," J. Psycho/., 1942, 


92. K. Lewin, A Dynamic Theory of Personality. McGraw-Hill Book Co., 1935. 

93. K. Lewin, "Concept, method and reality in social science: social equilibria and 
social change," Hum. Relat., 1948, 1, 5-40. 

94. K. Lewin, T. Dembo, L. Festinger, and P. S. Sears, "Level of aspiration" in 
J. McV. Hunt (Ed.), Personality and Behavior Disorders (Vol. I). The Ronald 
Press Co., 1944, 333-78. 

95. J. F. Lincoln, Incentive Management. Cleveland: Lincoln Electric Co., 1951. 

96. R. Linton, The Cultural Background of Personality. Appleton-Century-Crofts, 
Inc., 1945. 

97. R. Linton, Acculturation in Seven American Indian Tribes. Appleton-Century- 
Crofts, Inc., 1940. 

97a. S. M. Lipset and R. Bendix, Social Status and Social Structure: A Re-examina- 
tion of Data and Interpretations (Institute of Industrial Relations, Reprint 
No. 35). University of California, 1952. 

98. R. S. Lynd, Knowledge for What? The Place of Social Sciences in American 
Culture. Princeton University Press, 1939. 

99. R. S. and H. M. Lynd, Middletown, A Study in Contemporary American Culture. 
Harcourt, Brace & Co., 1929. 

100. R. S. and H. M. Lynd, Middletown in Transition. Harcourt, Brace & Co., 1937. 

101. D. W. MacKinnon, "Violation of prohibitions" in H. A. Murray, Explorations 
in Personality. Oxford University Press, 1938, 491-501. 

102. D. C. McClelland and F. S. Apicella, "A functional classification of verbal re- 
actions to experimentally induced failure," /. Abn. Soc. Psycho/., 1945, 40, 376- 

103. W. McDougall, An Introduction to Social Psychology, John W. Luce & Co., 

104. W. McDougall, Energies of Men. Charles Scribner's Sons, 1913. 

105. W. McGehee, "Judgment and level of aspiration," /. Genet. Psycho/., 1940, 22, 

106. R. B. McLeod, "The place of phenomonological analysis in social psychological 
theory" in J. H. Rohrer and M. Sherif (Eds,), Social Psychology at the Cross- 
roads. Harper & Brothers, 1951, 215-41. 

107. N. R. F. Maier, Psychology in Industry. Houghton Mifflin Co., 1946. 

108. N. R. F. Maier, Frustration. McGraw-Hill Book Co., 1949. 

109. R. Malinowski, A Scientific Theory of Culture and Other Essays. University of 
North Carolina Press, 1944. 

110. K. Marx and F. Engels, Manifesto of the Communist Party. Chicago: C. H. Kerr 
& Co., 1912. 

111. E. Mayo, "Day-dreaming and output in a spinning mill," /. Nat'l. Inst. Ind. 
Psychol. 1925, 2, 203-09. 

112. E. Mayo, The Human Problems of an Industrial Civilization. Harvard Business 
School, Division of Research, 1946. (First printing by The Macmillan Co., 1933.) 

113. E. Mayo, The Social Problems of an Industrial Civilization. Harvard University 
Press, 1945. 

114. G. H. Mead, Mind, Self and Society from the Standpoint of a Social Behaviorist. 
University of Chicago Press, 1934. 

115. M. Mead, Coming of Age in Samoa, William Morrow & Co., 1928. 


116. M. Mead (Ed.), Cooperation and Competition Among Primitive Peoples. 
McGraw-Hill Book Co., 1947. 

117. R. K. Merton, "The machine, the worker and the engineer/' Science, 1947, 
105-79, 84. 

118. D. C. Miller and W. H. Form, Industrial Sociology, Harper & Brothers, 1951. 

119. N. E. Miller, "The frustration-aggression hypothesis," Psychol. Rev., 1941, 48, 

120. G. Murchison (Ed.), A Handbook of Social Psychology. Clark University Press, 

121. G. Murphy, L. M. Murphy and T. Newcomb, Experimental Social Psychology. 
Harper & Brothers, 1937 (Revised Edition). 

122. T. M. Newcomb, Personality and Social Change. Dryden Press, 1943. 

123. T. M. Newcomb, Social Psychology. New York: The Dryden Press, Inc., 1950. 

124. T. M. Newcomb, E. L. Hartley (Eds.), Readings in Social Psychology. Henry 
Holt & Co., 1947. 

124a. T. Parsons, Essays in Sociological Theory: Pure and Applied. Glencoe, 111.: Free 
Press, 1949. 

125. "The people of the United States a self portrait," Fortune, (February) 1940, 
21 ff. 

125a. H. W. Pfantz, "The current literature on social stratification: Critique and 
bibliography," Amer. J. Sodol., 1953, 58, 391-418. 

126. L. Postman, J. S. Bruner, and E. McGinnies, "Personal values as selective fac- 
tors in perception," /. Abn. Soc. Psychol, 1948, 43, 142-154. 

127. L. Postman and G. Murphy, "The factor of attitude in associative memory," 
/. Exper. Psychol^ 1943, 33, 228-238. 

128. M. G. Preston, A. Spiers, and J. Trasoff, "On certain conditions controlling the 
realism and irrealism of expectations," /. Exper. Psychol., 1947, 37, 48-58. 

129. M. G. Preston and J. A. Payton, "Differential effect of a social variable upon 
three levels of aspiration," /. Exper. Psychol, 1941, 29, 351-69. 

130. F. J. Roethlisberger, Management and Morale. Harvard University Press, 1941. 

131. F. J. Roethlisberger, and W. J. Dickson, Management and the Worker. Harvard 
University Press, 1946, Chapter 13. 

132. C. R. Rogers, Counseling and Psychotherapy. Houghton Mifflin Co., 1942. 

133. J. H. Rohrer and M. Sherif (Eds.), Social Psychology at the Crossroads. Harper 
and Brothers, 1951. 

134. S. Rosenzweig, "Frustration as an experimental problem: VI A general outline 
of frustration," Character Person., 1938, 7, 151-60. 

135. S. Rosenzweig, "The experimental measurement of types of reaction to frus- 
tration" in H. A. Murray, Explorations in Personality. Oxford University Press, 
1938, 585-99. See also "An outline of frustration theory" in J. McV. Hunt 
(Ed.), Personality and the Behavior Disorders. The Ronald Press Co., 1944, 

136. E. H. Ross, Social Psychology. The Macmillan Co., 1908. 

137. E. Q. Rundquist and R. F. Sletto, Personality in the Depression. University of 
Minnesota Press, 1936. 

138. S. S. Sargent, "Conceptions of role and ego in contemporary psychology" in 
J. H. Rohrer and M. Sherif (Eds.), Social Psychology at the Crossroads. Harper 
& Brothers, 1951, 355-70. 

139. P. S. Sears, "Level of aspiration in successful and unsuccessful children," /. Abn. 
Soc. Psychol, 1940, 35, 498-536. 


140. R. R. Sears, Survey of Objective Studies of Psychoanalytic Concepts (Bulletin 
No. 51 ) . New York: Social Science Research Council, 1943. 

141. R. R. Sears, "Non-aggressive reactions to frustration/' Psychol Rev., 1941, 48, 

142. R. S. Seashore, and D. Katz, "An operational definition and classification of 
mental mechanism/' Psychol. Rev., 1937, 44, 1-24. 

143. R. Shafer and G. Murphy, 'The role of autism in a visual figure-ground relation- 
ship/' /. Exper. Psychol., 1943, 32, 335-343. 

144. M. Sherif, An Outline of Social Psychology. New York: Harper & Brothers, 1948. 

145. M. Sherif, "A Study of Some Social Factors in Perception," Arcft. Psychol., No. 
187, 1935. 

146. M. Sherif, The Psychology of Social Norms. Harper & Brothers, 1936. 

147. M. Sherif, "Introduction/' in J. H. Rohrer and M. Sherif (Eds.), Social Psychol- 
ogy at the Crossroads. Harper and Brothers, 1951. 

148. M. Sherif and R. Cantril, The Psychology of Ego-Involvements. John Wfley & 
Sons, Inc., 1947. 

149. R. Stagner, Psychology of Personality. McGraw-Hill Book Co., 1937. 

150. W. I. Thomas, "The province of social psychology/' Amer. J. SocioL, 1904, 10, 

151. W. F. Vaughn, Social Psychology. New York: The Odyssey Press, 1948. 

152. Wage Incentive Plans (Studies in Personnel Policy No. 68) . New York: National 
Industrial Conference Board, 1948. 

153. W. L. Warner and others, Social Class in America. Chicago: Science Research 
Associates, 1949. 

1 54. W. L. Warner and J. O. Low, The Social System of the Modern Factory. Yale 
University Press, 1947. 

155. H. C. Warren, Dictionary of Psychology. Houghton Mifflin Co., 1934. 

156. M. Wertheimer (Selected Papers), in W. D. Ellis, A Sourcebook of Gestalt Psy- 
chology. Harcourt, Brace & Co., 1938. 

1 57. J. West, Pldnville, US.A. Columbia University Press, 1945. 

158. W. F. Whyte, "Small groups and large organizations" in J. H. Rohrer and M. 
Sherif (Eds.), Social Psychology at the Crossroads. Harper & Brothers, 1951, 


159. W. F. Whyte (Ed.), Industry in Society. McGraw-Hill Book Co., 1946. 

160. W. F. Whyte, Human Relations in the Restaurant Business. McGraw-Hill Book 
Co., 1948. 

161. K. Young, Social Psychology. F. S. Crofts & Co., 1944. 

162. B. Zawadski and P. Lazarsfeld, 'The psychological consequences of unemploy- 
ment/' /. Soc. Psycho/., 1936, 6, 224-51. 

Field Theory and Action 
Research in Group Dynamics 

AMONG systematic contributions from experimental psy- 
chology which have had a strong influence upon the investigation of 
motivation and morale in industry are developments in the way of 
field theory and action research in group dynamics. 

Field theory* as formulated by Lewin, pictures psychological events 
as the outcome of dynamic forces operating within a mathematical- 
spatial construct constituting the individual's psychological field. A com- 
mon industrial situation can be used to illustrate elements of field theory 
(22, 23, 31) and also Lewin's contribution to motivational theory: 

"Let us suppose that a worker learns from the bulletin board that a 
job which he desires is open. This job is located in a company plant 
which is farther from his home than the office in which he is now work- 
ing. In Lewin's terminology, the desire for the job means that a region 
under tension exists within the worker. When he perceives that a new 
opening is available, he can be expected to make a move toward ob- 
taining it. Such locomotion could be anticipated so long as the person 
concerned is not restrained by impassable barriers or repelled by op- 
posing forces. In this situation, the job would be said to have positive 
valence for the person. Any object or condition which has positive 
valence attracts tne person from all points in the field, and is tnus the 
center of a force field in which forces are directed toward the object or 
condition from all points in the field/' 

1 The backgrounds of field theory are to be found, in part, in laboratory research 
conducted by Wertheimer (42, 43), Kohler (17, 18), and Koffka (16). They called 
attention to the significance of the total configuration or the Gestalt in psychological 
activities, and initiated a series of experiments on whole-part relationships which 
veered completely away from the atomistic approach of earlier studies of psychological 
processes and behavior. 



In considering the above psychological situation, field theorists would 
refer to the life space of the individual. As used by Lewin, life space is 
the "totality of facts which determine the behavior of an individual at a 
certain moment. The life space represents a totality of possible events. 
The life space includes the person and the (psychological) environment" 
(23, p. 216) of the person. 

Field theory treatment* of individual (and also social) psychology 
proceeds by making as accurate a description or diagnosis of the essentials 
of particular psychological situations as possible. Two kinds of concepts 
are used in these descriptions: spatial or topological and dynamic or 
vector concepts. Lewin believed that with the aid of certain arbitrary 
conventions it was possible to display both simple and complex psycho- 
logical fields in diagrams of two, or sometimes three dimensions, and to 
deduce from these diagrams the behavior to be expected of the individual 
or group. 

A life-space diagram which embodies some of Lewin's most commonly 
used concepts is presented in Figure 6. P represents a person, and the rest 

Figure 6. P = a person. A, B, C, and D are regions in the person's environ- 
ment. D is a goal region having ' 'positive valence" ( -f ) . The arrow represents 
the tendency of the person to move from A to D. C is a region which offers 
resistance to the person's passage (a "barrier").* 

of the diagram represents a situation in which the person is found at a 
particular moment. This situation has four regions: A (where the person 
now is), B, C, and D. One of these regions, D, attracts the person or 
constitutes a positive god. In Lewin's language, this region is said to have 
positive valence, and this fact is indicated by the plus sign. Whenever 
a positive valence is present, the person will tend to move toward the 
goal. This fact is represented by the arrow which points toward this 
region. The arrow symbolizes what Lewin calls a psychological force 

* The author is indebted to F. W. Irwin, University of Pennsylvania, for help in 
the simplified treatment of basic concepts in field theory on pages 110-112. 

* Acknowledgment is made to F. W. Irwin, University of Pennsylvania, for the 
above diagram. 


acting upon the person. The strength of this force can be indicated by 
the length of the arrow. 

The topologicd aspects of this situation are, then, the fact that the 
person is now in region A, and that, in order to reach the goal region, 
D, he must leave region A and go through regions B and C in that order. 
The dynamic aspects of the situation are the facts that region D has posi- 
tive valence and that a force of a certain strength is impelling the per- 
son toward that region. In addition, the diagram represents region C 
as having the dynamic characteristics of a barrier, i.e., of offering resist- 
ance to the person's progress through it; in the diagram, this is indicated 
by the cross-hatching. Another way of representing a barrier is to diagram 
it with a heavy lin^ as the entrance boundary of a region. Such barriers are 
thought of as having different degrees of resistance. 

Since the diagram shown in Figure 6 represents a set of abstractions, 
such as regions, barriers, valences, and forces, it can be employed for any 
psychological situation whose essentials are the same as those represented 
here. For example, the person may be the worker (P) referred to on 
page 110, who desires a transfer to a new job which can be obtained by 
going to the Personnel Department (B); arranging to use his own car 
for transportation (C), and completing the transfer to the new job (D) . 
It is assumed that transportation involves some difficulties or problems 
and therefore constitutes a "barrier/' However, the diagram might 
equally well represent a salesman who wants to interview a prospect (D) , 
and who must not only go to another part of the town (B) in order to 
reach him, but must get by the defenses of his prospect's secretary (C). 

From the same point of view, Lewin is able to represent situations in 
which the person is repelled from some region. Such a region is said to 
have negative valence, and is marked by a minus sign. The person's tend- 
ency to move away from such a region would be diagrammed by an ar- 
row pointing away from it. A conflict exists whenever the person has 
forces acting on him in opposing directions. In the example given earlier, 
such a conflict would exist if the new job wanted by the worker involved 
shift work which he considered to be undesirable. A similar conflict would 
exist if the salesman were somewhat intimidated as a result of an earlier 
unhappy experience with gruff ness on the part of his prospect. To diagram 
such conflict situations' characteristics would require the placing of a 
minus sign together with the plus sign in region D and drawing an arrow 
pointing away from region D with its head applied to the person, P. 

Those given above represent basic concepts or constructs in Lewin's 
motivational theory (33). In this theory, a state of tension, existing 
within one or more regions of the person, is "coordinated with those pat- 
terns of behavior that are generally recognized [in the more conventional 



treatment of motivation discussed in Chapter 4] as representing drives or 
motives; and the discharge of a system under tension as coordinated with 
the satisfaction of a drive or motive" (11, p. 206). However, these forces 
may vary with other factors, such as distance, cognition of barriers, etc. 

As is pointed out in later sections of this chapter, field theory, originally 
directed toward dealing with the psychology of the individual, has been 
embodied into social psychology. A further consideration of one of the 
concepts treated above, and of a number of additional constructs, is 
required as an introduction to the consideration of field theory in social 

The satisfaction of drives or motives involves movement toward or 
away from an object or situation. Such movement is described by Lewin 
in the term locomotion. As stated earlier, the direction of locomotion, 
i.e. toward or away from an object, will depend upon the strength of 
the attracting forces (positive valence) or/and of repelling forces (nega- 
tive valence) surrounding it. In the examples cited on earlier pages, loco- 
motion may be physical in character, as in the case of the worker who 
goes to the Personnel Department to arrange a transfer to the new job. 
However, inherent in Lewin's system is the view that locomotion need 
not involve actual movement in a physical world. 

Locomotion, in the Lewinian sense, may be "quasi-physical, quasi- 
social, and quasi-conceptual in nature" (23, p. 59) . Thus "to 'come closer' 
to another person through conversation" is a case of "social locomotion 
which, although it involves no physical movement, is psychologically 
real . . . There can be psychologically real locomotion in quasi-concep- 
tual fields" as in proceeding from an "unclear, unstructured region" in 
starting to solve a mathematical problem to a goal in the way of a solu- 
tion of the problem (23, p. 49). In illustrating such aspects of the con- 
struct "locomotion," Lewin uses the example of a 16-year-old boy (P, 
Figure 7) whose goal is to become a physician. 

Figure 7. Situation of a Boy Who Wants to Become a Physician.* P = per- 
son; G = goal; ce = college-entrance examinations; c = college; m = medi- 
cal school; i = internship; pr = establishing a practice. 

* By permission, from Principles of Topologicd Psychology, by K. Lewin. Copyright 
19 36, McGraw-Hill Book Company, Inc. 


The 'path' to this goal (G) leads through definite states: college- 
entrance examinations (ce), college (c), medical school (m), intern- 
ship (i), establishing a practice (pr). The boy may have a fairly clear 
idea of college. Medical school and the following stages may constitute 
a more or less undifferentiated region 'beyond' which lies the goal of 
being a physician. Of this the boy may have a false but nevertheless a 
clear picture. 

"When he passes his college entrance examinations he has made a 
'step forward' on the way to his goal. This movement is certainly not a 
bodily one. Nevertheless it is real locomotion, a real change of position 
in the quasi-social (and as a matter of fact also in the objective social) 
life space. The examinations have brought him a step closer to his goal. 
The reality of the change in his position becomes clear when one con- 
siders that many things are now within his reach which were not before. 
He can go to college or university, his time is much more within his 
own control than before. His social position too is changed: he can 
play on the college football team, go to the dances, etc. His examina- 
tions therefore had for him the character of a boundary between two 
distinct regions. He had to cross this boundary if he wished to go from 
the one region to the other. 

"Had he failed in his examinations, then he would not have made 
this advance toward his goal. But also in that case there would have 
been a real change in his life situation. The failure would have changed 
the barrier between him and the region of college, which was shortly 
before in his immediate neighborhood. The barrier would seem much 
more solid, almost impassable. The youth would be thrown back and 
possibly would seek an entirely new goal" (23, p. 48). 

In Lewin's system the life space of a person may correspond to the real 
world with which he is in contact. On the other hand, it may be out of 
touch with reality in the case of the person "living in a world of fiction 
and phantasy" (9, p. 217) . These possible differences in the life space are 
treated by Lewin as falling along a separate dimension which he calls 
the dimension of reality-unreality. "A day dream, a vague hope, has in 
general less reality than an action; an action sometimes has more reality 
than speech; a perception more than a change; a faraway 'ideal' is less real 
than a 'real goal' that determines one's immediate action. Action itself/' 
adds Lewin, "can be of very different degrees of reality" (23, p. 196). 
Thus, processes involving strong needs which can be overcome only by 
surmounting strong barriers have a high degree of reality. However, in 
the face of such needs, the person may resort to "wishful thinking," rep- 
resenting a high degree of irreality. 

In later developments of his system Lewin added still another dimen- 
sion to the life space, that of time perspective* This, as will appear later, 
is thought to have particular implications for production and morale 

A concept first used by Frank. Sec L. K, Frank, "Time perspectives " 7. Soc. 
^ 1939, 4, 293-312. * * ' 


in industry. Time perspective, as defined by Leeper, is "the definition 
by the person, at any given moment, of what has been the pattern of his 
relevant psychological past, and of what probably will be his future situa- 
tion and position with regard to some other regions of valences which 
are now important in his life space" (20). In other words, time perspec- 
tive includes not only a person's present outlook but, in addition, his past 
and future as they affect him at that moment. Thus, in the example of 
the worker interested in transfer to a new job, an evaluation of his past 
experience and of the potentials of the new job would constitute part 
of the worker's present life situation. 

Action Research in Group Dynamics 

As stated earlier, the field theoretical approach was first applied by 
Lewin to problems of individual behavior. The construct of psychologi- 
cal field was originally developed as an aid in dealing with the develop- 
ment of individual personality. 

Although the first systematic application of field theory to social prob- 
lems is found in the work of Brown (4), the construct of social field, 
pointing toward a field theoretical study of group life, is nevertheless 
implicit in the early work of Lewin. This "social" viewpoint found frui- 
tion in action research and in research in group dynamics (24, 27). 

Action research is a method of field research oriented toward collect- 
ing objective information to be used as a basis for action decisions. Action 
research was viewed by Lewin as consisting of experiments conducted in 
the actual social situation, with key individuals in the group involved 
being drawn in as co-operating recorders, observers, and analyzers. 

Information collected for action decision is also analyzed from the 
theoretical point of view by research personnel interested in theory of 
group dynamics. As implied in the term, group dynamics places emphasis 
upon interpersonal relations social interactions of individuals in groups 
and among groups (41 ), the term "dynamic" referring to the forces which 
facilitate or inhibit changes in groups. 

The special significance of field theory for social psychology is that it 
looks upon the individual not as a static entity or unit, but rather as a 
dynamic being whose characteristics and behavior change under the vary- 
ing influences of the social field of force of other persons and events con- 
stituting the whole situation. This leads directly into the field theoretical 
study of group life. According to Lewin, Lippit, and Escalona, "A field 
theoretical study of group life must: ( 1 ) regard the group as a whole, ex- 
isting in a larger social field, with many overlapping dynamic relationships 
(influence of family and school membership, etc.); (2) regard the group 
as composed of interdependent parts of members (implies criterion of 


interdependence as basic definition of a group); (3) regard each mem- 
ber as existing in a social field in which even the 'individual' problems 
must be viewed in a framework of group membership; (4) demands that 
observational analysis shall recognize these basic assumptions by ob- 
serving group and individual activities in terms of meaningful wholes, 
always relating units of observation to their larger setting" (25, p. 53) . 

Industrial Implications of Field Theory 

The "impact" of field theory on industrial psychology is to be found 
in (a) increased concern for the forces in the total industrial situation 
acting upon the individual and (b) an emphasis upon the analysis of 
interactions within groups and among groups in the industrial setting. 
Since good or bad industrial relations depend upon the adjustment of an 
aggregate of individuals, the problem in industrial relations is to dis- 
cover the strains and stresses, the internal and the external pressures 
which force the point of equilibrium outside of the worker's area of 
tolerance to the point where the total situation becomes unbearable 
to him (8). An objective in this approach is to help set standards in 
the environment so as to make it improbable that conditions will strain 
the powers of adjustment of a significant number of employees. In terms 
of field theory, this cannot be done without considering (a) the indi- 
vidual worker as a whole personality, subject to various internal tend- 
encies in the total environment which make various demands upon him 
and (b) the relations among competing and integrating forces in the 
total situation (10). 


Industrial applications of field theory were initiated, under the direction 
of Lewin, in the plant of the Harwood Manufacturing Company. Re- 
search in this area and also in the application of field theory to other 
social problems was extended through the organization of a Research 
Center for Group Dynamics, originally located at the Massachusetts In- 
stitute of Technology, and later transferred to the campus of the Uni- 
versity of Michigan (27). This was conceived as a research center for 
building an experimental social science which would draw upon and 
co-ordinate the fields of psychology, sociology and cultural anthropology 


Parallel with the development of the Research Center for Group 
Dynamics in this country, there arose in England, at the close of World 
War II, the Tavistock Institute of Human Relations, described by Jaques 


as an "outpatient clinic for social disorders/' which has applied field 
theoretical and psychoanalytic concepts and techniques, and also socio- 
metric procedures, in the study of social interactions (14, 46). As in the 
American center, attention was focused upon industrial morale and other 
problems in industry, as well as upon field theoretical research in other 
social situations (13). Research in industry is represented by a 3-year 
investigation conducted in the plants of the Glacier Metal Company 
(12, 45). Here, a field theory approach was used (1) to study the psy- 
chological and social forces affecting group life, morale, and productivity 
of a single industrial community; (2) to formulate more effective ways 
of resolving tension within and between groups and to find out how to 
overcome resistance to change. Research directed toward these ends was 
conducted by teams of professional workers simultaneously responsible 
for both making research observations and providing professional advice 
to the firm. According to the investigators, this approach, taken from 
clinical medicine, brings the team "into contact with areas of the life 
of the factory which in normal circumstances remain inaccessible" (12, 
p. 15) to research personnel and provides access to data which might 
otherwise remain unrevealed (13a). 

The resulting case history of developments "in the social life of one 
industrial community" furnishes information as to the variety and com- 
plexity of forces involved in the social behavior of factory personnel and 
insight in field theory terminology with respect to the magnitude, 
distance, direction, strength and valence of such forces. Although not 
discussed in detail in this volume, the survey is of interest in spite of the 
failure to use measurement techniques in providing a "description of 
the field of forces, at a first level of approximation" (12, p. xv), operating 
over a period of time within the total situation represented by a "going" 
business concern. 


The impact of field theory context in the United States is seen most 
clearly in investigations at the level of the "face-to-face sub-group" (7a), 
described in later chapters. 4 Thus, in the Harwood Manufacturing Com- 
pany investigation, described on pages 164-169, labor turnover and 
absenteeism, viewed as indicators of low morale, are ascribed to frustra- 
tion associated with conflict between ( 1 ) increasing difficulty in meeting 
a production "goal" as a worker nears it, and (2) an associated increasing 
desire or drive to do so ( 36, p. 33) . Lowered morale is thus viewed as the 
end effect of a discrepancy between "level of aspiration" and actual per- 
formance. Applying field-theory constructs, resistance to change to a 

4 See particularly Chapter 9. 


new job appearing in a slow rate of relearning and below-standard out- 
put is referred to negative attitudes associated with "locomotion" from 
production "regions" with strong "positive valences" to "regions" of 
strong "negative valences" in terms of pay, job security, factory status, 
and personal success (37). 

Participation in Decision-Making: As appears in Chapter 9, employee 
participation in decision-making plays a large role in Lewin's approach 
to the problem of motivation in industry. In a number of experiments, 
Lewin and his coworkers have shown that production can be increased 
by creating a social situation in which workers are, in Allport's terms, 
"participant in cooperative activity" ( 1 ) . Employee participation in de- 
cision-making, within the scope of the field theoretical approach, be- 
comes an effective .device for lowering resistance to change, and raising 
production, by lowering the resistance of barriers to the "goal" of higher 
output (3, 6, 38). While increased production can be obtained by 
pressure which increases the strength of the forces pushing in the direc- 
tion of the desired change, the use of group participation permits smoother 
"locomotion" to the same "goal" without the creation of "tensions" 
which may lead to industrial strife. Participation in decision-making in 
industry is generally viewed as an experience wherein attitudes favorable 
to change are taken on by the workers. The group-determined decision, 
according to Lewin, provides the link between motivation and action 
(32, p. 233). 

Group decisions are also said to depend partly on how the "goal" and 
the total situation are viewed by the group. As a result, before a deci- 
sion to change can be made, there must be a change in group per- 
ception which includes also a perception of the result of change. Prior 
to effective group action on a proposed change, the objective must be 
clarified; the way to the "goal" and the available means for reaching this 
"goal" must be determined. A strategy of action must precede action. 
These, according to Lewin, are among the feedback problems of social 
diagnosis and action (28). 

As has been stated in Chapter 5, frustration arises when the individual 
is faced by goals with different "valences" or even by goals associated 
with "forces" of different strength (32). In the opinion of Lewin and 
his coworkers, participation in decision-making contributes to the avoid- 
ance of such conflict situations. The potential for industrial strife is low- 
ered, since the change in group perception associated with group par- 
ticipation tends to bring the production "goal" closer to the standard 
desired by management. Furthermore, "emotionality" is lowered, since 
workers playing the "role" of planners tend to keep discussion at a rela- 
tively depersonalized level ( 30) . 


As indicated on page 114, "time perspective" is viewed by Lewin as 
playing a crucial role in morale. A "positive time perspective" is not only 
a basic element of good morale, but "at the same time/' according to 
Lewin, "the process is reciprocal; high morale itself creates a long-range 
time perspective and sets up worthwhile goals" (26, p. 64). The "life 
space" of every person includes not alone the present, but the past and 
future as well, as they affect him at the moment. The feelings, emotions, 
actions and morale of a person at any given time depend upon his total 
"time perspective." Thus, members of a factory group may be expected 
to reach higher levels of production and have better morale if there is a 
realistic relation between their actual production and "level of aspiration" 
for future production. This is illustrated, according to Lewin, by the fact 
that a group of new employees for whom a production goal was set 
weekly progressed more rapidly, and reached higher levels of produc- 
tion, than did a group whose members were simply told, after a week's 
training, that they were producing only 20 to 25 per cent of the standard 
set for skilled operators (26), 

Authoritarian and Democratic Leadership: Of special interest to in- 
dustrial psychology is the impetus given by Lewin and his associates to 
controlled experiments with social groups which include an analysis of the 
leader and the effects of authoritarian, democratic and laissez-faire types 
of leadership upon the group and its individual members (24, 34). 
According to Bavelas, who has applied field theory to industrial re- 
search, motivation and morale reach high levels in a "democratic" at- 
mosphere where there is active participation between the leader and 
members of his group (2, 3). In the "autocratic" group the work goals 
are set by the leader; it is he who determines work morale and makes 
the group an organized unit. In contrast to this, the democratic group 
is more "we-centered"; its members play an active part in setting group 
goals in a "permissive" atmosphere. As a result, the "time perspective" 
of members of the group is more independent of the leader than is the 
case of the "autocratic" group. As a further consequence, it can be antici- 
pated that the production and morale of the "democratic" group will 
"hold up" better than that of the autocratic group, should either lose 
its leader (27). 

The Social Climate: The contrast between authoritarian and demo- 
cratic social climate has served as a point of departure for "field theoreti- 
cal excursions" into the processes and social resultants of group life (25). 
Techniques for experimentally inducing "group atmospheres" originally 
employed with children in laboratory situations have been adapted for 
the investigation of autocratic, democratic, paternalistic, and laissez-faire 
atmospheres in industry (34). Attention has been directed toward the 


problem of communication, since inadequate communication as well as 
poor leadership become sources of frustration both in the formulation 
of "goals'' and in the effective achievement of individual and of group 
"goals" (15, 35). 

The study of the nature of leadership and of varieties of "social cli- 
mates" has led directly to field theoretical investigations bearing upon 
the structure of the group. Thus, from a study of work satisfaction in the 
textile industry, Hutte concludes that "... an industrial organization 
is a whole system of relations, in which the parts in a complicated way 
influence each other in their own functioning. The properties of this 
'whole' are not observable by studying the properties of the constituent 
parts ... the parts function hierarchally, allowing some parts to have 
a more embracing pr more centralized influence within the whole. This 
hierarchal functioning, may be denoted by the concept of integration" 
(10, p. 190). According to Hutte, a field theoretical context is required 
in research on group structures to understand this phenomenon of in- 

Changing Group Structures: Principles for inducing changes in groups, 
of interest from the viewpoint of motivation and morale in industry, 
have been derived from application of the field theoretical approach by 
Festinger, Schacter, and Back (7). Their studies of how groups function 
and why they function in a particular way have suggested that homo- 
geneity, physical proximity, and group norms are important variables re- 
lated to group functions. The effectiveness of group norms or standards, 
it is said, is associated with cohesiveness of the group, and the presence 
of subgroups tends to disrupt the cohesiveness of the "parent" group. 
Group norms exhibit "power fields" or spheres of influence in relation 
to both morale and group action. Social groupings, it was found, create 
definite channels of communication, and the type of information most 
frequently communicated through these channels is that relevant to the 
functioning of the group. A further significant outcome of these studies 
is the conclusion (representing a confirmation of Lewin's view) that it 
is easier to change a whole group at once than to change an individual 
while leaving the remainder of the group unchanged 


As will become evident from the detailed discussion of studies in later 
chapters, it is not always apparent that the results achieved by investi- 
gators employing field theoretical contexts can be accounted for only by 
reference to field-theory constructs. It seems true, as suggested by Krech, 
that "Kurt Lewin and his co-workers . . . have made brilliant and lasting 
contributions to the formulation of psychological laws and techniques 


of analysis where the individual's 5 psychological field was the subject of 
study" (19, p. 678). Furthermore, Lewin and his associates unquestiona- 
bly deserve considerable credit for undertaking to provide a compre- 
hensive theoretical context and for enlarging the techniques of action 
research in the investigation of group life and of social atmospheres.* 
Lewin's insistence that without a "good" theoretical foundation applied 
research follows a path of trial and error, and becomes misdirected and 
inefficient, has had a profound influence upon research bearing upon 
motivation and morale in industry. Findings from individual studies have 
helped to establish the importance of the view (earlier formulated by 
the Hawthorne investigators) that the levels of motivation and morale 
are not necessarily the end effects of specific incentives of specific ele- 
ments in the work situation or specific items of personnel policy 
(39, 40, 44) but of the whole or total situation and of the many over- 
lapping dynamic interrelations of its parts, involving both the individual 
and the smaller groups in a larger social field (25). 

Acknowledgment of this debt to field theory does not call for the be- 
littlement of theoretical and factual contributions from other sources. 
From the viewpoint of theory, there may be justification for the ques- 
tion, raised by Krech, as to whether "genuine group law" can actually 
be established by merely transforming principles established for indi- 
vidual behavior by "substituting the word 'group' for 'individual' and 
the term 'social field' for 'psychological field' in the original formula- 
tions" (19, p. 678). Whether or not the specific criticism by Krech is 
justified, there nevertheless appears a need, in the current striving toward 
definitive laws of psychological reactions, for a balanced approach which 
calls for a tolerant view with respect to all possible explanations of social 
interactions. It is in this spirit of eclecticism that this section on theoreti- 
cal backgrounds has been written, and an effort made to interpret the 
findings of studies described in later chapters with reference to alterna- 
tive explanations, rather than upon the assumption that any single theory 
represents the "one bright hope in an otherwise desperate social pic- 
ture." T 

5 Not italicized in the original. 

It should however, be noted that this viewpoint of action research was anticipated 
in earlier studies conducted in the Hawthorne Plant of the Western Electric Com- 
pany (40, 44) under the influence of Mayo (39). 

7 A phrase used by Tolman (41) in referring to the concepts and techniques of 
field theory as elaborated by Lewin and his coworkers into a progress of "group dy- 



1. G. W. Allport, 'The psychology of participation," Psychol Rev., 1945, 52, 117- 

2. A. Bavelas, and K. Lewin, "Training in democratic leadership," /. Abn. Soc. 
Psychol, 1942, 37, 115-19. 

3. A. Bavelas, "Some problems of organizational change," /. Soc. Issues, 1948, 3 

(Summer), 48-52. 

4. J. F. Brown, Psychology and the Social Order. McGraw-Hill Book Co., 1936. 

5. E. D. Chappie (with C. M. Arensberg), "Measuring Human Relations," Genet. 
Psychol. Monog., No. 22, 1940. 

6. L. Coch, and J. R. P. French, Jr., "Overcoming resistance to change," Hum. 
Relat., 1948, 1, 5U-33. 

7. L. Festinger, S. Schacter and K. Back, Social Pressure in Informal Groups. Harper 
& Brothers, 1950. 

7a. J. R. P. French, Jr. and A. Zander, "The group dynamics approach," in A. Korn- 
hauser (Ed.), Psychology of Labor-Management Relations. Champaign, 111.: In- 
dustrial Relations Research Assn., 1949, 71-80. 

8. J. M. Fraser, "A psycho-dynamic approach to industrial relations," Occup. Psy- 
cho/., 1946, 20, 133. 

9. E. R. Hilgard, Theories of Learning. Appleton-Century-Crofts, Inc., 1948. 

10. A. A. Hutte, "Experiences in studying social psychological structures in industry," 
Hum. Relat., 1949, 2, 185-192. 

11. F. W. Irwin, "Motivation" in H. Helson (Ed.), Theoretical Foundations of 
Psychology. D. Van Nostrand Co., Inc., 1951, 200-47. 

12. E. Jaques, The Changing Culture of a Factory. London: Tavistock Publications, 

13. E. Jaques, "Field theory and industrial psychology," Occup. Psychol^ 1948, 22, 

13a. E. Jaques, "Interpretive group discussion as a means of facilitating social change," 
Hum. Relat., 1948, 1, 533-49. 

14. E. Jaques, "Some principles of organization of a social therapeutic institution," 
/. Soc. Issues, 1947, 3 (Spring), 4-11. 

15. H. H. Kelley, "Communication in experimentally created hierarchies," Hum. 
Relat., 1951,4, 39-57. 

16. K. Koffka, Principles of Gestalt Psychology. Harcourt, Brace & Co., 1935. 

17. W. Kohler, Gestalt Psychology. Liveright, 1929. 

18. W. Kohler, Dynamics in Psychology. Liveright, 1940. 

19. D. Krech, "Psychological theory and social psychology" in H. Helson (Ed.), 
Theoretical Foundations of Psychology. D. Van Nostrand Co., Inc., 1951, 656- 

20. R. W. Leeper, Vector and Topolo&cal Psychology. University of Oregon, 1943. 

21. K. Lewin, Field Theory in Social Science (Edited by D. Cartwright). Harper 
& Brothers, 1951. 

22. K. Lewin, Dynamic Theory of Personality. McGraw-Hill Book Co., 1935. 

23. K. Lewin, Principles of Topological Psychology. McGraw-Hill Book Co., 1936. 


24. K. Lewin, "Field theory and experiment in social psychology: concepts and 
methods/' Amer. J. Soc., 1939, 44, 868-97. 

25. K. Lewin, R. Lippit, and S. K. Escalona, Studies in Topological and Vector Psy- 
chology. University of Iowa Studies, New Series No. 380, 1940. 

26. K. Lewin, 'Time perspective and morale" in G. Watson (Ed.), Civilian Morale. 
Houghton Mifflin Co., 1942, 48-71. 

27. K. Lewin, "The Research Center for Group Dynamics at M.I.T.," Sociometry, 
1945, 8, 126-36. 

28. K. Lewin, "Frontiers in group dynamics II," Hum. Relat, 1947, 1, 143-154. 

29. K. Lewin, Resolving Social Conflicts (Edited by G. W. Lewin and G. W. All- 
port). Harper & Brothers, 1948. 

30. K. Lewin, "The solution of a chronic conflict in a factory" in K. Lewin, Resolving 
Social Conflicts (Edited by G. W. Lewin and G. W. Allport). Harper & 
Brothers, 1948, 125-141. 

31. K. Lewin, "Nature of field theory" in N. Marx, Psychological Theory. The Mac- 
millanCo., 1951, 299-315. 

32. K. Lewin," Frontiers in group dynamics" in Field Theory in Social Science (Edited 
by D. Cartwright). Harper & Brothers, 1951, 188-237. 

33. K. Lewin, "Constructs in field theory" in Field Theory in Social Science (Edited 
by D. Cartwright). Harper & Brothers, 1951, 30-43. 

34. K. Lewin, R. Lippitt, and R. K. White, "Patterns of aggressive behavior in ex- 
perimentally created social climates," /. Soc. Psychol., 1939, 10, 271-99. 

35. R. Lippit, Training in Community Relations. Harper & Brothers, 1949. 

36. A. J. Marrow, "Kurt Lewin," /. Soc. Issues (Supplement Series No. 1), (Decem- 
ber) 1948, 27-34. 

37. A. J. Marrow, "Group dynamics in industry," Occupations, 1948, 26, 472-76. 

38. A. J. Marrow and J. R. P. French, Jr., "Changing a stereotype in industry," 
J. Soc. Issues, 1945, 1,3, 33-38. 

39. E. Mayo, The Human Problems of an Industrial Civilization. Boston: Harvard 
Business School, Division of Research, 1946. (First printing by The Macmillan 
Co., 1933.) 

40. F. J. Roethlisberger and W. Dickson, Management and the Worker. Harvard 
University Press, 1939. 

41. E. C. Tolman, "Kurt Lewin 1890-1947," /. Soc. Issues (Supplement Series I), 

(December) 1948, 22-27. 

42. M. Wertheimer, "Untersuchungen zur Lehre von der Gestalt, I and II," Psycho/. 
Forsch., 1922, 1, 47-65; 1923, 4, 301-50. 

43. M. Wertheimer (Selected Papers) in W. D. Ellis, A Source Book of Gestalt Psy- 
chology. Harcourt, Brace & Co., 1938. 

44. T. N. Whitehead, The Industrial Worker, Harvard University Press, 1938. 

45. A. T. M. Wilson, "Some aspects of social process," /. Soc. Issues (Supplement 
Series No. 5), (November) 1951. 

46. A. T. M. Wilson, "Some implications of medical practice and social casework for 
action research," /. Soc. Issues, 1947, 3, No. 2. 


The Influence of Job Preference, 
Size of Work Group, and 
Knowledge of Results 

MOTIVATIONAL theory, discussed in Part II, provides a 
frame of reference for research directed toward the identification and 
evaluation of significant motives, attitudes, and incentives in industry. 
Progress toward the maximum utilization of human resources in increas- 
ing productivity and in raising the levels of job satisfaction and morale 
calls for fact-finding investigations conducted in industrial plants and 
through associated laboratory studies. The prime object of such investi- 
gations will be that of determining how the worker reacts to factors in 
and outside of the plant which may influence his feelings, sentiments, 
attitudes, and behavior in the plant. 

The scope of investigation in the area of motivation and morale is al- 
most unlimited. It calls for an intensive study of the influence of hours 
of work, methods of wage payment, plant physical conditions, personnel 
policies and practices, structural organization of the company, labor- 
management relations, and also of the effects of age, schooling, marital 
status, socio-economic status, etc., upon employee attitude and per- 
formance. A comprehensive research program will involve a detailed and 
direct study of the feelings, attitudes, and sentiments of individual work- 
ers by means of such scientific techniques as are available. It will lead 
to a comparison of different groups in the same plant and of different 
plants as an aid in the isolation of the factors which determine the at- 
titudes of the individual employee and the morale of workers as a group 
or "class" in the structuring of society as a whole (15). It will invariably 



and emphatically stress the observation of facts and their interrelations 
as the sine qua non of the scientific study of motives-in-work. 


Of the various research methods available for the investigation of the 
sources of motivation and morale, "true experiments" and accidental 
"natural experiments" generally yield the most direct and unambiguous 
results (16). In the former, specific and controlled changes are made 
in the work situation and data are accumulated on parallel changes in 
workers' feelings, sentiments, attitudes, and performance. Such experi- 
ments are typified in the study of wage incentive systems, by Wyatt, 
discussed below, and in the Hawthorne experiment, described in Chap- 
ters 10 and 11. In the second type of investigation, advantage is taken 
of existing or newly introduced variations in the work situation and com- 
parisons made of the productivity, satisfaction or morale of groups ex- 
posed to specified differences in objective conditions. Illustrative of such 
"natural experiments" are the comparison of output of differently sized 
work groups by Marriott, described below, and studies of the morale of 
"low-production" and "high-production" groups and comparisons of pro- 
duction and morale under different supervisors, conducted by the Uni- 
versity of Michigan Survey Research Center, discussed in Chapter 8. 

The number of both "true" and "natural" or "ready-made experi- 
ments" is small, in part because of the hesitation of industry to permit 
research. For example, 66 replies from 158 member firms surveyed by the 
backwardness on the part of industry in supporting "human relations" 
research. For example, 66 replies from 158 member firms surveyed by the 
American Management Association, in 1949, showed that only "a dozen 
companies are doing work of a type which can properly be described as 
personnel research, even though they are spending on it only a penny 
or two out of each personnel function dollar . . . Two-thirds of the 
companies replying reported that they are doing no such research at all, 
and a few wrote that they were too busily engaged in collective bargain- 
ing to reply" ( 5 ) . At the same time, according to the survey report, "prac- 
tically every reply to the AMA survey began or ended with an expression 
of gratitude that a start was being made in presenting what management 
is doing to raise the knowledge of human behavior in industry" (5). 

In spite of general lack of management support, a number of com- 
panies have provided the opportunity for "true" and "natural" experi- 
ments. Such research, typified in experiments described in Part III of this 
volume, has been supported not only by industry, but with funds pro- 
vided by government agencies, research foundations, and universities, 


supplemented frequently by contributions of the time and efforts of 
university personnel. 

An Investigation of the Role of Financial 
and Nonfinancial Incentives 

Illustrative of such experiments is an investigation, sponsored by the 
Industrial Health Research Board, which also provides a striking example 
of the possibility of studying simultaneously both the effectiveness of 
different wage-incentive plans and the influence upon production of fac- 
tors other than the desire for more money. The significance of this study 
is particularly marked because of the fact, referred to in Chapter 2, that 
wage-incentive systems are seldom installed under conditions which 
permit an evaluation of different wage-incentive plans and of the direct 
effect of the wage incentive in comparison with the influence of other 

The subjects of the experiment were ten girls aged between 15 and 16 
years, selected on the basis of personal characteristics as shown during 
an interview and in the performance of special tests (24). The experi- 
ment was conducted in a chocolate manufacturing plant, and the ma- 
terial and methods of work employed were similar to those used in the 
corresponding departments in the factory. The operatives were part of 
the ordinary factory staff, but the experiment was conducted in a separate 
room specially equipped for the purpose by management. 

Under ordinary conditions each worker is usually employed on a single 
operation. In the present experiment each of the ten workers learned 
and became thoroughly experienced in five different repetitive opera- 
tions. These, in order of increasing complexity as evidenced in average 
time taken to complete the operation, included (1) iwwrcjppwg, (2) 
wrap/ring, (3) packing, (4) weighing, and (5) 'weighing and wrapping. 

The hours of work were from 8 to 12 and from 1 to 5, but the first 5 
and the last 10 minutes in each spell were devoted to answering questions 
on feelings of fitness, boredom, fatigue, and the like, making the actual 
number of hours worked in each spell 3%. Output was recorded every 
quarter hour. In addition, the behavior of the operatives during work, 
their remarks and appearance, were noted by the experimenter. 


In accordance with the usual practice in industry, the workers in this 
experiment were first paid a fixed weekly wage (Time Rate) . This method 
of payment was continued for 9 weeks and was followed by a competi- 



tive-bonus-wage system which remained in operation for 15 weeks (Bonus 
Rate). During the latter period the operatives were ranked each week 
according to output, and the slowest worker was given the same wage as 
before. The next in order of output received an additional sixpence and 
so on throughout the group, so that the fastest worker earned 4s. 6d. 
more than the slowest. This system of payment was followed by a flat 
Piece Rate which remained unchanged for another 12 weeks. Thus, for 
a period of 36 weeks, the different methods of payment represented the 
only significant external changes in the conditions of work. 

As a first step in the analysis of results, the production records of all 
the workers engaged in the five different processes were combined, pro- 
duction during successive weeks being expressed as a percentage of the 
output during the .first week of the experiment. As is apparent from 
Figure 8, the amount of improvement from week to week during the 

I 2 3 4 

IS 20 


26 30 35 

Figure 8. The Influence of Various Wage Systems upon Production* 

Time-Rate period was relatively small, averaging 12 per cent. There were 
also indications that a continuation of this method of payment would 
have resulted in no appreciable increase in output. The immediate effect 
of the introduction of the Bonus Rate was an increase in output of 46 
per cent, and the output curves continued to rise for the greater part 
of this period, becoming stabilized at a fairly uniform level by the end of 
the 15 weeks. Following the introduction of the flat Piece Rate there 
was an almost immediate increase in output of 30 per cent which, after 

* By permission, from S. Wyatt, Incentives in Repetitive Work, (Industrial Health 
Research Board Report No. 69), Her Britannic Majesty's Stationery Office, 1934. 


a temporary drop, was maintained with very little change throughout the 
remainder of the 12-week period. 

In a later phase of the investigation, the Piece Rate was retained in 
the wrapping and packing operations, the Bonus Rate was substituted 
in the weighing and weighing and wrapping processes, while in unwrap- 
ping the original Time Rate was revived. The most significant outcome 
of such changes was the marked decrease in output caused by the re- 
vival of the Time Rate. The findings suggest that the lower output caused 
by the time rate is not confined to the early stages of the learning process, 
but may be equally apparent in the case of more experienced workers. 
This aspect of the situation was further investigated by substituting a 
Piece Rate for the Time Rate in the case of a group of experienced work- 
ers employed in feeding wrapping machines with chocolate. The intro- 
duction of the wage incentive system resulted in an average increase in 
output of 36 per cent. As in the case of the earlier experiment reported 
by Kitson (14), it would appear that in the absence of an adequate 
financial incentive, these experienced workers had settled down to a 
production rate well below the limits of their potential achievement 

It is evident that, for the type of work done in both the main experi- 
ment and subsequent studies, the Time Rate proved to be a weak in- 
centive to production. Under this method of payment output became 
stabilized at a level much below the limits of possible achievement. An 
analysis of observations made by the experimenter and of interview 
findings showed that work under this system of wage payment was char- 
acterized by an attitude of indifference on the part of the operators 
concerned, and that output "seemed to be largely dependent on constant 
supervision and the fear of dismissal" (24, p. 55). The Time Rate was 
conducive to boredom and responsible for an increase in the number 
and duration of voluntary stoppages although, on the other hand, it 
tended to reduce instances of jealousy among employees with different 
production rates and to promote friendly relations. 1 

The introduction of the wage-incentive systems resulted not only in 
marked increases in production, but also in increased interest, in de- 
creases in the amount of lost time, and in fewer instances of "trouble- 
some" conduct. However, particularly under the Bonus Rate, there was 
an increase in the frequency of disagreements among workers and of 
other signs of stress including, under the Piece Rate, greater faultfinding 
with materials and with other conditions which seemed to interfere 
with productive activity. The experimental design did not permit a clear- 
cut evaluation of the comparative merits of these two wage-incentive 

* See Chapter 11. 


systems since, as shown in a supplementary study, the relative increases 
noted under these systems might have been a function of the order in 
which they were introduced. However, according to the investigators, 
the Piece Rate was regarded by the workers as the fairest of the three 
methods because individual rewards were dependent upon individual 
achievement and because it tended to produce closer application as evi- 
denced, for example, in decreased talkativeness, fewer instances of trouble- 
some conduct, and a reduction in the frequency and duration of stop- 


The analysis of the results from all the processes studied leads to the 
conclusions that "the rate of working both during and after the learning 
period is largely dependent on the method of payment" and that "out- 
put seems to become stabilized at a level which is proportional to the 
strength of the monetary incentive" 2 (24, p. 9). However, a further 
analysis of the results, applying to individual processes, also provides evi- 
dence that, in a work situation which fails to satisfy other basic needs or 
wants, wage incentives will not, in themselves, produce that "prolonga- 
tion of effort" (18, p. 52) which, according to Mace, is among the prin- 
cipal beneficial effects of incentive systems. 


Evidence as to ineffectiveness of wage incentives under such condi- 
tions appears in a comparison of the preferences for the five work 
processes included in the main experiment by Wyatt and his associates. 
From weekly statements obtained from the 10 workers it was found that 
the order of preference for the five operations was as follows 

L Wrapping 

2. Packing 

3. Weighing and wrapping 

4. Weighing 

5. Unwrapping 

Figure 9 shows the relative output in the 36 successive weeks of the 
main experiment for each of the five work processes, each point repre- 
senting the combined results of ten workers Expressed as a percentage 
of output of the first week. It can be seen that the effect of the wage 
incentives is most marked in the operations which aroused the more 
favorable feeling and that it is completely absent in the case of the 

* Not italicized in the original. 












Figure 9. Relative Output in Successive Weeks in Different Processes under 
the Influence of Various Wage-Payment Plans * 

least popular .operation, unwrapping. As a matter of fact, in the most 
preferred operation, wrapping, the rate of production was trebled by 
the end of the experiment, while in the very similar but converse 
process of unwrapping there was practically no improvement, and out- 
put throughout the experimental period was usually less than that re- 
corded during the first week. An analysis of observations of the experi- 
menter and of interviews suggested that the lack of improvement in the 
latter operation could be attributed, in part, to the belief of the workers 
that unwrapping was a wasteful and comparatively useless procedure. 
Neither continued practice on the task nor wage incentives produced any 
effect in the way of stimulating output on an operation to which the 
worker attached little value which "failed to give that satisfaction 
which accompanies the production of attractive and saleable commodi- 
ties" * (24, p. 26). Such findings support the viewpoint, which is gaining 

8 Since the processes differed in complexity, it could be hypothesized that the rate 
and amount of improvement in any process is a function of the complexity and of 
the number of times each cycle of movements is repeated in a given period. Accord- 
ing to the investigators, an examination of the results showed that the possible effects 
of such factors seem to be submerged by the influence of the extent of interest in the 

* By permission, from S. Wyatt, Incentives in Repetitive Worfc, (Industrial Health 
Research Board Report No. 69), Her Britannic Majesty's Stationery Office, 1934, 


ever widening acceptance, that a financial incentive may have no effect if 
the -work operation is disliked; if the work jails to satisfy or conflicts with 
other sentiments or needs which influence the will-to-work. 


Other results from the main experiment in the series of studies by 
Wyatt and his associates provide clues as to other characteristics of the 
work situation which affect the will-to-work (17, 18). These showed that 
variation in the rate of working of different individuals was dependent 
upon the prevailing group or social atmosphere^ i.e., on a "collective in- 
fluence which formed an intangible background and determined the 
general nature of the reactions to the conditions of work." However, 
certain individuals Were found to have a particularly great influence on 
the output, attitude and behavior of the remaining workers. For exam- 
ple, the absence of such a worker, characterized as being of the "domineer- 
ing and disturbing" type caused an increase in output of 12.5 per cent 
on the part of the remainder of the group. The production of an opera- 
tive was also found to be closely related to that of her immediate neigh- 
bor. The output of a worker tended to increase when she was moved 
from the neighborhood of a slower worker to that of a faster worker and 
vice versa. In several such cases the change in output varied from 15 to 
20 per cent and there were striking changes in the disposition, moods 
and pleasure accompanying work. The average output of individual 
workers increased when each worker was isolated, in turn, from the 
remainder of the group. Such findings, according to the investigators, 
"show the interactions between one worker and another and illustrate 
the futility of regarding the individual as an independent unit . . . 
When/' they add, "workers are employed as a fairly compact group they 
no longer behave as individual units but respond as a collective whole 
to the prevailing group 'atmosphere' . . . The creation of a suitable 
attitude or 'atmosphere' is accordingly of some importance and as a 
rule there is scope for improvement in this direction" (24, p. 51 ) . 

The Size of the Work Group 

The observation that workers employed as a compact group react as a 
"collective whole" raises a question concerning the relation between the 
size of the work group, on the one hand, and production, job satisfac- 
tion, and morale, on the other. A survey of 22 American plants, reported 
by Balderston (1), in 1930, showed that 16 firms preferred work groups 
consisting of fewer than 50 individuals, and that 8 of these considered 
20 to be the optimal upper limit. A study reported by Handyside (7a), 
in 1952, indicates that the average size of the primary working group in 


British industry is 21.9, and that the average number in groups of prod- 
uction workers varies between 16.8 (in chemical plants) and 33.3 (in 
textile plants). Not alone Balderston, but also Dickinson (7), follow- 
ing a study of wage systems, reaches the conclusion that "small groups 
provide the greatest incentive although larger groups are practical in 
progressive manufacture where automatic conveyors or other transport- 
ing and reporting devices may enable workers to co-operate who cannot 
see or hear each other at work" (21, p. 47) . 


The relation between the size of work group and output has been 
investigated, by Marriott (21), in two Britsh motorcar factories. Em- 
ployees in both plants were men on direct production work who were 
paid on the basis of total earnings of the group in which they worked, 
the amount being divided pro rata to the base rate of each individual 
and the number of hours he worked. In Factory A, the wage system was 
in the form of a group bonus, based on an "efficiency" percentage, i.e., 
the ratio of the time allowed for an operation to the time actually taken, 
expressed as a percentage. The efficiency percentage was used as a meas- 
ure of group output in Factory A. In Factory B, payment was based on 
group piecework, each group being paid on the basis of the number of 
units produced. In Factory B, the average piecework earnings (in pence) 
per man per hour were used as a measure of group output. Both measures 
are directly proportional to output, and can therefore be used in com- 
paring the productivity of work groups. 

In Factory A, efficiency percentages were collected for 15 consecutive 
months (five quarterly periods), from January, 1946, to March, 1947. 
Throughout this period, the groups were very stable as regards size, and 
the mobility of individual men was normal for existing industrial condi- 
tions. In Factory B, there were marked variations in over-all conditions 
during the period of the experiment. For example, in order to counter- 
balance production needs, some groups were varied in size from time 
to time and there was some movement of the men involved. Because 
of these and other factors, collection of data was limited to the four most 
stable six-week periods, free from holidays and which, although differ- 
ing in the number and distribution of hours worked, provided full-time 
and similar working conditions for all the groups. However, interpreta- 
tion of earnings data was complicated in Factory B by the fact that a 
temporary additional bonus had been allowed on many of the assembly 
lines, in the form of an addition of from 7% to 12% per cent to the 
usual earnings, varying according to the production planned. Since this 
bonus was not proportional to output, it was deducted from the group 



earnings before the average net piecework earnings (the output measure 
used) were calculated. 

Elimination of 16 groups for which data were incomplete in Factory A 
left a total of 153 groups for analysis in this plant, distributed with re- 
spect to size as shown in Table 8. In Factory B, of the total groups 51 
were constant in size throughout the 18 months; 20 varied in size, gen- 
erally moving up or down one class only in one or two periods; 16 groups 
were present in only three of the four six-week periods used in the 
analysis, and 18 in only two. In the treatment of the individual periods, 
the latter were included, two appearances being considered as a reasona- 
ble criterion in defining the existence of a group. As shown in Table 8, 

TABLE 8. Number and Size of Groups Studied in Investigating the Relation 
Between Size of Working Group and Output (after Marriott) 

Size of 
group num- 
ber of 

Factory A 

Factory B 

number in 

of groups 

in group 

Number of groups Periods 





Under 10 
50 and over 









All groups 




86 98 86 79 

even this treatment provides too few of the larger sized groups, especially 
30-39 and 40-49, to give "reliable" averages, although it should be re- 
called that data were averaged for six-week periods and for fairly large 
numbers of men. 


The over-all relation between group size and output (as represented 
in earnings under incentive wage systems) is shown in Figure 10, Part A. 
Differences in output among variously sized groups were found to be 
substantial in practical terms, ranging from ah average efficiency per- 
centage of 126.1 to one of 133.8 in Factory A and from 31.0 to 36.4 in 
hourly piecework earnings (in pence) in Factory B. When the average 
output of the smallest group is given a value of 100, the averages of 
the smallest groups prove to be from 6 to 8 per cent higher than those of 


ti z 



Part A: Output. 

O C9 

4 1 

s < 





Part B: Percentage of Men "Without Knowledge/' 











Part C: Level of Satisfaction with Payment System of Men 

"Without Knowledge/' 


Under \O-\9 20*29 


Size OF 

Figure 10. Relations among Size of Work Group, Knowledge of Results, 
Output, and Satisfaction.* 

* By permission, from H. Campbell, "Group incentive payment schemes: the effects 
of lack of understanding and of group size/' Occup. PsychoL, 1952, 26, 15-21. 


groups containing between 40 and 49 men * (22). Correlations between 
size of group and output in each of the quarterly and six-week periods, 
respectively, were all found to be negative and statistically significant, 5 
ranging from .203 to .277 for all groups in Factory A, and from 
.287 to .491 for all groups in Factory B. 

Both plants were located on the outskirts of towns of 100,000 popula- 
tion, and travel, housing and similar facilities were much alike. The 
nature of the work processes was almost identical in the two factories, 
and physical conditions and welfare amenities were of a high order and 
much alike. However, there were marked differences between the two 
plants in terms of employee satisfaction, the quality of supervision, em- 
ployee co-operation in meeting quotas, personnel policies, etc. The fact 
that "in these two contrasting conditions, one in which the workers were 
stimulated to increase productivity, and the other in which restriction 
appeared to be forced on them, the effects of group size were so con- 
sistently similar*' (21, p. 55), is interpreted as showing that group size 
influences productivity. 


Interviews with managers and foremen in the two factories showed 
that their preference was for snialler-sized groups. 6 In Factory A, 75 per 
cent expressed opinions favoring groups smaller than 50, the majority 
preference being for groups with from 20 to 40 men. Similar results were 
obtained in Factory B. Workers' opinions are not reported in the publi- 
cation on this study. However, investigations conducted in the United 
States have provided some information concerning the relation between 
group size and the feelings and sentiments of both employees and of 
young people planning to enter the labor market. Thus, from studies con- 
ducted by the Survey Research Center, University of Michigan, Katz 
concludes that "employees of small work groups are more satisfied than 
employees of large work groups" (12, p. 168). Evidence as to pref- 
erences of young people with respect to size of work group was obtained 

* In both factories, the output of groups with 50 or more men was higher than 
for groups with 40 to 49 men. It is indicated that it was not easy to explain this 
tendency for output to rise in the groups with 50 or more men. It is suggested that 
the more frequent use of power conveyors by such large groups may account for the 
situation. According to Marriott, "it seems likely that the more constant rate of work- 
ing imposed by them (i.e., the mechanical conveyors) counterbalanced the effect of 

special allowances. 

any of these studies, is suggested in 
the findings of an exploratory study by Hemphill (8) which suggest that "the size 
of the group is a variable which to some degree conditions leader behavior." 


in a survey of 6,789 tenth- and twelfth-grade students in 56 Michigan 
public and private high schools, conducted by the Social Research Serv- 
ice, Michigan State College in 1949. As shown in Figure 11, young people 
prefer to work in small work groups (26). Studies in a metals-fabricating 
plant, by Kerr, Koppelmeier and Sullivan (13) lead to the conclusion 
that "large departments, even though they may encourage many aspects 
of job satisfaction, perhaps fail to provide as attractive long-term goals 
as do small departments" (13, p. 51). 

Relationships between group size and morale have been noted in a 
study, by Hewitt and Parfit, conducted in a British textile plant (8a). 
Here, the "non-sickness" absence rate was found to be significantly greater 

It I were employed by a concern f would prefer to work: 

Entirely by myself 1 4% 

Mostly by myself but 
with a few contacts 
with other employees 

With o small group 
of workers (under 25) 

As a member of a 
large group of workers 




No response I 4% 

Figure 11. Attitudes of High-School Children toward Size of Work Group * 

among workers in large-sized rooms than among those in smaller rooms 
accommodating fewer employees. On the assumption that a high ab- 
sence rate is an indicator of low morale, it is hypothesized that "there 
may be some analogy between the spread of poor morale and the spread 
of infection/' According to the investigators, "one would expect the 
risk of coming into contact with poor morale, like the risk of infection, 
to be greater in the larger workshops" (8a, p. 41). In some respects, the 
findings of this British study are in agreement with those from an earlier 
study in the aircraft industry, conducted during World War II by Mayo 

* By permission, from Youth and the World of Work, Michigan State College, 


and Lombard (23), which suggested that absenteeism (thought of as 
an objective index of employee morale) 7 was a function of group team- 
work. "Natural," congenial work groups, conducive to such teamwork 
and associated advantages, are said to develop more readily when the 
number in the group is small, although strong cohesion may also occur in 
larger "family" groups characterized by a substantial core of employees 
who have worked together for a long time. 


In considering the findings of the study in the British motorcar fac- 
tories, Marriott also reaches the conclusion that "the inverse relation- 
ship between group size and productivity is the result of greater co- 
hesion in small teams because men know each other better, can see each 
other at work and, consequently, are less suspicious and require less super- 
vision" (22, p. 18). However, exception is taken to the viewpoint, in- 
herent in Mayo's theory of "informal social organization" in industry, 8 
that size of work group is important only when groups develop "nat- 
urally," and not when groups are organized as working units by man- 
agement. On the contrary, according to Marriott, "as far as output is 
concerned the size of management-organized groups is important . . . 
the study of the optimum size for each situation may lead to greater 
efficiency, and the resulting higher earnings to greater satisfaction" 
(21, p. 57). 

As has been suggested by Marriott, further research is needed partic- 
ularly in the form of controlled or "true" experiments to determine 
optimal size of work groups for different types of work and under various 
organizational conditions. Nevertheless, available data suggest the need 
for management attention to the question of the size of the work unit. 
Consideration of this problem will involve not alone the question of 
number, but also the need for partitions and other surroundings which 
will help establish a "small group atmosphere." Management must be 
prepared not only to keep the work group small, but also to cast aside 
its predilection for the "wide open spaces" in offices and factories, and 
to provide physical settings which will contribute to the integration of 
small work groups into productive and satisfied social units. 

The Influence of "Knowledge of Results" in 
Relation to the Size of the Work Group 

Although group cohesion appears to be in itself an important factor 
in accounting for higher output in smaller work groups, other elements 

'See pages 211-213. 
See Chapter 11. 


in the situation may contribute to this result. Campbell has suggested 
that "the worker's knowledge of the relation between his effort and earn- 
ings is likely to decrease as the size of the group increases, and there may 
be a corresponding weakening of the incentive to work" * (4, p. 15). 


This hypothesis has been tested through an analysis of data obtained 
in the two British motorcar factories, referred to in preceding pages, 
in which Marriott investigated the relation between size of the work 
group and output. In both factories the "usual" methods for explaining 
incentive schemes were used, including booklets, explanations to indi- 
vidual workers, and detailed instructions for supervisors and for workers' 
representatives. Nevertheless, interviews with a 10-per-cent sample of 
workers in Factory A showed that two-thirds of the 340 men working 
under the efficiency percentage plan had no effective knowledge of re- 
sults of their efforts, as represented by their incentive earnings, until they 
received their wage sheets in the following week. In Factory B, even 
with a much simpler payment system (group piecework) , one-quarter 
of the 217 men (20-per-cent sample) interviewed were in a similar 

A further analysis indicated a definite tendency, evident in Figure 10, 
Part B, for the proportion of workers without knowledge of results to 
become larger as the size of the work group increased. Of further interest 
in Figure 10, Part B, is the fact that in Factory B the proportion of work- 
ers without knowledge of results did not rise in groups including 50 to 
99 men. According to Campbell, the workers in such groups were as- 
sembling whole cars, and it was therefore quite easy for them to count 
the number going off the line. This effect did not occur in groups of 100 
or more men also engaged in assembling cars in Factory A, apparently 
because the larger size of the group made the counting more difficult. 


Interviews with workers yielded an "index of satisfaction' 1 toward 
the wage-payment system, expressed on a scale ranging from 2 or "very 
satisfied'' to 2 or "very dissatisfied." As shown in Figure 10, Part C, 
with the exception of one group in Factory A, the level of satisfaction 
of the workers without knowledge of results decreased as the size of the 
group increased. On the other hand, group size was found to have no 
effect on the satisfaction of those with knowledge. In order to check 

9 Not italicized in the original. 


these findings, derived from separate analyses in the two factories, the 
workers from each factory were combined into two groups representing 
satisfied and dissatisfied workers. In the case of those without knowledge 
of results, the proportion of dissatisfied workers became progressively and 
significantly larger as the group size increased (X 2 = 19.44; n=5; 
P = < .01), but there was no consistent or significant difference in the 
case of those with knowledge (X 2 = 4.36; n = 4; P = > .30) . 

According to Campbell, the explanation of these findings may be 
found partly in the social structure of the groups. The compactness of 
the small group allowed its members "to get to know each other well," 
and information (including presumably that concerning the incentive 
system and output) could be passed on quickly. In general, a common 
purpose appeared t6 bind together the members of the smaller groups 
"evoking loyalty to the group and a trust in each other." In contrast, the 
sustaining "quality of belonging to a known and trusted social unit" 
appeared to be absent in larger groups, with resulting adverse attitudes 
in the way of suspicions of malingering on the part of fellow workers, 
dislike of dependence on others, etc. However the findings also clearly 
indicate that knowledge of results operates in determining feelings and 
attitudes under wage incentive systems. 


The emphasis in the experiment described immediately above was on 
knowledge of results in relation to the size of the work group. However, it 
brings into relief the role of knowledge of results, sometimes known as 
K of R, as an independent and direct incentive to production and satis- 
faction at work. Many laboratory experiments and a number of indus- 
trial and military studies have shown that efficiency of learning can be 
increased by supplying knowledge of results, in the way of information 
on progress in learning (25). 

Illustrative of experiments on the motivating influence of knowledge 
of results is a series of studies conducted by the Applied Psychology Unit 
of Cambridge University (20) . In one of these, reported by Macpherson, 
Dees, and Grindley (19), the subject was given the task of pressing a 
telegraph key for a set time of approximately 0.7 seconds. Under one 
set of conditions subjects were able to discover h,ow well they were doing 
by observing a scale on a mirror galvanometer. Under the second set of 
conditions they were deprived of this knowledge of results. Figure 12 
shows the average results of 10 subjects on a series of runs involving the 
alternation of 10 trials with knowledge of results (K) and trials with- 
out knowledge of results (N), followed by another series of runs with 
knowledge of results (K). Section A of Figure 12 shows the progress 






uj 10 


t t 

t t 1 1 t 

t t 

i i t t 

10 i 


O With knowledge of results 

10 I 10 

X Without knowledge of results 

Figure 12. Average Errors of Ten Subjects with Alternate Rims of K and 
N Trials.* Curve A represents the average errors during the very first run 
of ten K trials. Curve B represents the average errors during the whole subse- 
quent fifteen N runs. Curve C represents the average errors during the whole 
subsequent fourteen K runs. 

* By permission, from S. }. Macpherson, V. Dees, and G. C, Grindley, "The effect 
of knowledge of results on learning and performance/' Quart. /. Exper. PsychoL, 1948, 
1, 68-78. 


of learning when knowledge of results is given. Section B indicates clearly 
the marked and statistically significant deterioration which takes place 
when knowledge of results is withdrawn, while Section C shows the im- 
provement when the subject is again given an opportunity to observe how 
well he is getting along on the task. 

The principle that the learner makes most rapid progress when he is 
aware of how well he is doing and of the nature and number of errors he 
is making has been put to practical use in a number of military situations. 
For example, in the case of radio-code instruction, the traditional method 
of training was to present a whole series of signals before the trainee was 
given information as to whether or not his identifications of signals were 
right or wrong. Studies conducted during World War II resulted in the 
introduction of a "code-voice" method, in which the correct letter was 
given immediately after the trainee had written down the signal as he 
understood it (9, 10). The use of this method, which gave immediate 
knowledge of results (and also included the feature of "whole-learning") 
resulted in reducing by more than 25 per cent the time taken to reach a 
speed of five words per minute, and in decreasing the proportion of those 
failing to reach this speed from 15 to 3.4 per cent. In another military 
situation, it has been shown that men learned to track a gun more 
accurately when the instructor sounded a buzzer to indicate that the 
tracker was off target than when they were not immediately informed of 
these errors (2). Conversely, according to Lindsley, men learning radar 
operation became less accurate, in the course of six days of practice, be- 
cause they were not given information as to their errors and rate of 
progress (11). 


Studies of the influence of knowledge of results as an incentive have 
been almost exclusively limited to learning situations. The possibility that 
K of R can have useful implications for stimulating production in the 
day-to-day activities of experienced workers is illustrated in an experience 
reported by Bingham in 1932. Following an analysis, conducted by Outh- 
waite, of the fireman's job in an electric generating plant, it was de- 
cided that, in addition to what the firemen know as to the efficiency of 
each battery of six boilers, data were needed as to the actual operation of 
the separate boilers stoked by different men. As a result, instruments 
were installed to provide each man with information as to what was 
expected of his own boiler and precisely how efficiently or inefficiently 
it was working. "For the first time," according to Bingham, "the men 
really knew when they were being successful in their work. A new spirit 
spread among them. A fresh pride replaced the old indifference. Graphic 


daily and weekly records for each boiler were posted. Rivalry sprang up, 
competition between the three eight-hour shifts. Moreover, each man 
could compare his own performance today with his last week's record, 
and try to improve it" (3, p. 262). 

The experiment was so successful that it was extended to other batteries 
of boilers and later to all the power plants. Manuals of instructions were 
prepared for use by the men in learning how to manage their furnaces still 
more economically. The firemen were given a new status, in the form of a 
new job title of "stoker operator/' Because of the increased efficiency in 
operation, it was found possible to grant a substantial raise in wages. 
Labor turnover dropped. The company profited in the form of savings in 
cost of coal, which was estimated to be $330,000 a year in the main plant 
alone. But, says Bingham, "the psychologist chiefly values the fact that 
(as a result of changes premised upon knowledge of results) the men now 
liked their work. They had a sense of accomplishment. Their work had 
become worth while" (3, p. 263). 

Keeping Workers Informed of Progress 
and Achievement 

Management and supervisory personnel in American industry have not 
been unaware of the need for keeping workers informed of their accom- 
plishments. Many supervisors and foremen who are unfamiliar with the 
phrase "knowledge of results" make a practice of supplying workers with 
an answer to the question "How'm I doing?" on a day-to-day basis. 
Numerous devices have been developed to help keep workers informed 
about the rate and quality of their work. Nevertheless, there are also in- 
dications of a failure to make advantageous use of the fact that groups of 
workers without knowledge of results tend to show a persistently lower 
level of performance than workers on the same tasks who continuously 
know how well they are doing. Perhaps further consideration should 
be given, particularly in repetitive work, to wider use of readily visible 
output-recording devices which, according to Wyatt, Langdon and Stock, 
would not only provide K of R but would "induce" the operator to com- 
pete against himself (24). As a matter of fact, it is suggested by these 
British investigators that the incentive would be particularly strong if the 
indicator registered the accumulating output in terms of pounds and 
shillings (or dollars, in the U.S.A.). 

As is evident from the discussion of restriction of output in Chapters 2 
and 3, accurate and immediate records of output can produce "boomer- 
ang" effects where plant conditions or policies and practices are other- 
wise adverse to the stimulation of the will-to-work. In particular, work- 
ers become suspicious and frustrated when they are unable to understand 


the wage systems under which they work (6, p. 144), and are therefore 
unable to determine how they are doing. Under no circumstances can 
there be any justification for a wage-incentive system which is so unclear 
either in structure or in its presentation that workers can derive from 
it no knowledge of their accomplishments from day to day. Simplifica- 
tion of the structure and functioning of wage systems, and clear explana- 
tions of their operation to workers, can help eliminate such difficulties. 
Steps in this direction are important because, as clearly shown in this 
chapter, knowledge of results, obtained through an understanding of 
the wage system, enhances the value of wages as an incentive to produc- 
tion and contributes to satisfaction and adjustment on the job. 

In Summary: As shown in this chapter, "true" and "natural" experi- 
ments are throwing light upon the value of particular incentives and 
specific sources of motivation in industry. Such experiments confirm the 
viewpoint that the financial incentive, while important, is not the sole 
medium for stimulating production, increasing job satisfaction, and rais- 
ing the level of employee morale. Both production and satisfaction tend 
to be higher in jobs that are "liked" than on those that are "not liked." 
The size of the work group affects output and attitudes, which both tend 
to be better in smaller-sized groups. Immediate knowledge of results acts 
as a stimulating influence, although its effects are also related to a pattern 
of interpersonal relation which is apparently more characteristic of smaller 
than of larger groups. It is apparent, from the studies cited, that effects 
of specific incentives, such as those considered in this chapter, can be 
isolated. However, there also appears a consistent thread of influence 
derived from the group make-up which requires consideration, along with 
and also apart from discrete and measureable factors such as "liking 
the job," size of the work group, and knowledge of results. 


1. C. C. Balderston, Group Incentives. University of Pennsylvania Press, 1930, 

2. W. C. Biel, G. E. Brown, R. M. Gottsdanker and R. C. Hall, The Effectiveness 
of a Check Sight Technique for Training 40 mm Gun Pointers (OSRD Report 
4054). Tufts College, 1944. 

3. W. V. Bingham, "Making work worth while" in Psychology Today. University 
of Chicago Press, 1932, 262-64. 

4. H. Campbell, "Group incentive payment schemes: the effects of lack of under- 
standing and of group size," Occup. Psychol., 1952, 26, 15-21. 

5. E. Dale, "Company research in human relations/' AMA Management News, 
June 23, 1949, 3-4. 

6. M. Dalton, "Economic incentives and human relations" in L. R. Tripp (Ed.), 


Industrial Productivity. Madison, Wis.: Industrial Relations Research Assn., 

7. Z. C. Dickinson, Compensating Industrial Effort. The Ronald Press Co., 1937, 

7a. J. D. Handyside, "An estimate of the size of primary working groups in British 
industry/' Occup. PsychoL, 1952, 26, 106-07. 

8. J. K. Hemphill, "Relations between the size of the group and the behavior of 
'superior' leaders," /. Soc. Psychol., 1950, 32 (First Half), 11-22. 

8a. D. Hewitt and J. Parfit, "A note on working morale and size of group," Occup. 
Psychol., 1953, 27, 38-42. 

9. F. S. Keller, "Studies in international Morse code. I., A new method of teach- 
ing code reception," /. Appl Psychol, 1943, 27, 407-415. 

10. F. S. Keller, The Radio Code Research Project: Find Report of Project SC-88 
(OSRD Report 5379). New York: The Psychological Corporation, 1945 (PEL 

11. D. B. Lindsley, Radar Operator Training: Results of Study of SCR-270-271 
Operators at Drew Field (OSRD Report 1737). Yerkes Laboratories of Primate 
Biology, 1943 (PEL 18367). 

12. D. Katz, "Morale and motivation in industry" in Current Trends in Industrial 
Psychology. University of Pittsburgh Press, 1949, 145-71. 

13. W. A. Kerr, G. J. Koppelmeier, and T. }. Sullivan, "Absenteeism, turnover and 
morale in a metals fabrication factory, ' Occup. Psychol., 1951, 25, 50-55. 

14. H. D. Kitson, "Extra wage incentive plans from a psychological viewpoint/' 
Bull. Amer. Mgt. Assn. 9 (Production Executive Series No. 9) 1925. 

15. A. Kornhauser, 'The motives-in-industry problem," Ann. Amer. Acad. Pol. Soc. 
Science, 1923, 110, 105-116. 

16. A. Kornhauser, "Psychological studies of employee attitudes," /. Consult. PsychoL, 
1944, 8, 127-43. 

17. C. A. Mace, "Satisfactions in work," Occup. Psychol, 1948, 22, 5-19. 

18. C. A. Mace, Incentives Some Experimental Studies (Industrial Health Research 
Board Report No. 72). H.M. Stationery Office, 1935. 

19. S. J. Macpherson, V. Dees, and G. C. Grindley, "The effect of knowledge of 
results on learning and performance," Quart. J. Exp. Psycho/., 1948, 1, 68-78. 

20. N. H. Mackworth, Researches on the Measurement of Human Performance 
(Medical Research Council Report No. 268). H.M. Stationery Office, 1950. 

21. R. Marriott, "Size of working group and output," Occup. PsychoL, 1949, 23, 

22. R. Marriott, "Socio-psychological factors in productivity," Occup. PsychoL, 1951, 
25, 15-24. 

23. E. Mayo and G. F. F. Lombard, Teamwork and Labor Turnover in the Aircraft 
Industry of Southern California. Harvard Graduate School of Business, Business 
Research Series No. 32, 1944. 

24. S. Wyatt (Assisted by L. Frost and G. L. Stock), Incentives in Repetitive Work 
(Industrial Health Research Board Report No. 69). H.M. Stationery Office, 

25. D. Wolfle, "Training" in S. S, Stevens (Ed.), Handbook of Experimental Psy 
chology. John Wiley & Sons, Inc., 1951, 1267-86. 

26. Youth and the World of Work. Michigan State College, 1949. 


Supervision, Productivity, 
and Morale 

CONSIDERATION of the problems of production, job satis- 
faction, and morale has been marked by a constantly recurring note em- 
phasizing the influence of supervision and of management. Thus, accord- 
ing to Elliott Smith, "A company pays the same for labor whether it is 
managed well or whether it is not, but what it gets from its labor depends 
not only upon the methods and equipment that the management pro- 
vides, but upon whether the employees work with a will. This in turn 
depends largely upon the skill of the junior executives in dealing with hu- 
man nature'' (18, p. 220). 

The Relation between Supervision and Output 

The hypothesis that motivation and morale are influenced by the 
quality of supervision has been tested in a number of experiments. Evi- 
dence that the supervisor can play an important role in stimulating 
output is found in an experiment reported by Feldman ( 1 ) . Among the 
employees of an insurance company were approximately 1,000 clerks, 
divided into 22 sections. In 1933, a new wage plan providing group in- 
centives for the 22 sections was inaugurated. Costs for each section for 
the previous 12 months were computed and each was allowed a group 
bonus on the savings it could effect over the cost for that period. All 
members of a section, including the supervisor, shared in these savings 
monthly on the basis of salaries, no change being made either in the basic 
salary or in established policies with respect to salary increases. For the 
year 1933, every section showed some improvement in performance, rang- 
ing from 2 per cent to 12 per cent, with an average of 8 per cent. 

In 1934, management effected a general shift of all section heads, in- 


volving a transfer of those who had been in charge of above-the-average 
bonus groups to below-the-average sections. One objective in this was 
to determine whether differences in results were related primarily to 
differences in supervision or to differences in the quality of personnel 
or conditions of work in the various sections. An analysis of production 
for the year 1934 showed increases in production in all sections, ranging 
from 6 per cent to 18 per cent. The order of merit of supervisors remained 
practically the same as in the study made at the end of 1933. In spite of 
reassignment to new sections, those section heads who stood high at the 
end of the first year were at the top of the list during the second year and 
vice versa. Changes in relative order were limited to three cases of super- 
visors who had moved a step or two. 

Early in 1935 the management again shifted, by lot, 20 of the 22 super- 
visors. Although they were reassigned by chance, supervisors generally 
maintained the same rank orders, in terms of improvements in the pro- 
ductivity of their groups, as they had held in the preceding two years. 

Characteristics, Attitudes, and Behavior 
of "High" and "Low" Production 
Supervisors in an Office Situation 

In the situation described above, the fact that employee performance 
improved in all sections with the introduction of the wage-incentive plan 
may be taken as evidence that the desire for more money motivated 
employees to better production. However, it is evident that the extent 
of improvement in any group was a function of still another incentive con- 
dition namely, the quality of supervision. In the study reported by Feld- 
man, apparently no effort was made to determine and define the char- 
acteristics, attitudes, and behavior of supervisory personnel which might 
contribute to variations in the motivation, productivity, and morale of 
employee groups. Information of this kind is to be found in a series of 
studies conducted as part of a long-range research program in human rela- 
tions and group organization being carried on by the Survey Research 
Center of the University of Michigan (2, 11). 

The initial investigation in this series, undertaken in 1947 with the 
support of the Office of Naval Research, was conducted in the home 
office (Newark, NJ.) of the Prudential Insurance Company (9) with 
(nonunionized) clerical workers and their supervisors as subjects. 

The experiment involved personnel from two departments. One of 
these, the Ordinary Policy Department, includes six parallel divisions, 
each duplicating the work of every other division in organization, type 
of work, type and number of personnel. Each division is made up of 11 


sections, each performing a specialized function, and each having a dupli- 
cate in every other division. In the second department, i.e., Debit Policy 
Department, there are four parallel divisions containing four parallel sec- 

Productivity in each of the units referred to above is measured by means 
of a budget system which provides a record of the ratio of actual clerical 
time spent in completing a given amount of work to an "expected" base. 
Essentially, the productivity measure is one which provides an index 
of the personnel costs for completing a given amount of work. The 
"budget ratio" or productivity index is characterized as a valid measure 
for comparing sections, and likewise divisions, doing the same kind of 
work during any period of time. 

Productivity differences between comparable sections were not found 
to be great rarely in excess of ten per cent. Nevertheless, the study was 
undertaken because some of the comparable sections showed significant 
differences in productivity over an eight-month period, even though the 
differences were small. 

The experiment involved 12 pairs of sections drawn from the two de- 
partments referred to above, each pair including sections with about the 
same number of people, at the same job levels, doing the same kind of 
work. Ten high-low productivity pdrs were selected for comparison from 
the Ordinary Policy Department (Department A) and two high-low pairs 
from the Debit Policy Department (Department B) * The pairs from De- 
partment A included sections which provided a range of work from highly 
routine typing to skilled calculation; of size of work group from 6 to 26; 
of types of employees from girls recently graduated from high school 
to men who had been with the company many years. The sections came 
from five of the six divisions, and sections which belonged in the same 
division were not consistently either the high or low sections in a produc- 
tivity pair. On the other hand, the high sections in Department B came 
from one division and the low sections from another. As a result, high- 
low section comparisons for Department B were at the same time division 
comparisons. Furthermore, the sections in Department B, averaging about 
32 women employees, were larger than the samplings from Department A. 

Since all sections and divisions were operating under the same work- 
ing conditions and the same general company policies, it seemed possi- 
ble, according to the investigators, to assume that there were only two 
variables which could account for group differences in productivity: (1) 
management and supervision within the sections and divisions, and (2) 

1 Sections used in the productivity analysis were only those which remained signif- 
icantly different from each other at the 5-per-cent level (or better) for the eight- 
month period ending September 30, 1947. 


interpersonal relations among employees in the work group, i.e. the "$o- 
cial psychology of the immediate work group! 9 There was accordingly 
set up an experimental design "with productivity as the dependent varia- 
ble, supervision and management as the independent variables and 
worker morale as the intervening variable" (6, p. 3). The design of the 
study employed three sets of independent measures, viz.: (1) a measure 
of productivity, (2) a measure of employees' perceptions and attitudes, 
and (3) a measure of supervisory perceptions, values, practices, and at- 

The measure of productivity, in the form of a "budget ratio," has been 
referred to above. The remaining measures were derived from free-answer 
interviews conducted with 24 first-line supervisors (section heads) and 
419 nonsupervisory, i.e. rank-and-file employees from the paired high 
and low sections. 2 Supervisory interview schedules included: (1) a de- 
scription of supervisory behavior, both in relation to subordinates and 
to superiors, and (2) attitudes toward their own jobs, subordinates, supe- 
riors, the company and company policies. The employee interviews meas- 
ured perceptions of and attitudes toward: (1) their own work groups, (2) 
their own jobs, (3) company and company policies, (4) job status and 
salary, and (5) supervision. 

Comparisons of high- and low-section heads showed striking similari- 
ties in amount of formal education and in length of service with the com- 
pany. There were no significant differences between the high and low 
supervisors in other demographic characteristics, such as age, sex, and 
marital status. However, analysis of the findings suggested that super- 
vision of men-productivity sections differed from those of "Low-produc- 
tivity sections in certain well-defined characteristics, attitudes, and be- 
havior. For example, as shown in Table 9, employee-centered supervisors 
proved to be higher producers (in terms of the productivity of their 
groups) than production-centered supervisors. 3 A supervisor was classi- 
fied as "employee-centered" when he showed a tendency to consider inter- 
est in his employees rather than in production as being of primary im- 

* The groups described in this section were only part of a larger number studied, 
which included 73 supervisory and managerial interviews and 742 nonsupervisory 
interviews. The latter are to be used in an analysis of the patterning and interrela- 
tionship of employee attitudes. 

8 Chi-square tests were made on the null hypothesis that there was no real differ- 
ence between supervisors or employees in the "highs" and those in the 'lows" on a 
given item or index. The null hypothesis was rejected and the relationship described 
as statistically significant when it was found that the obtained value would be likely 
to occur by chance five times or fewer in 100 if the true difference between "highs ' 
and "lows" were really zero. In such a case it was assumed that the obtained differ- 
ence was not due to chance alone. All tables reproduced in this section present results 
which are statistically significant at the 5-per-cent level or better. 


TABLE 9. Prudential Insurance Company: Relation of Section Heads 9 Atti- 
tudes toward Employees to Section Productivity (after Productivity, 
Supervision and Morale in an Office Situation, Survey Research 

Center, 1950) 

Employee- Production- Not 
centered centered ascertained N 

Heads of 

high sections 





Heads of 

low sections 





TABLE 10. Prudential Insurance Company: Relation of Proportion of Time 
Spent in Supervision to Section Productivity (after Productivity, Supervi- 
sion and Morale in an Office Situation, Survey Research Center, 1950) 

Question: "What proportion of your time is given to supervisory matters? 
What proportion to other duties?" 

50% or more Less than 50% 

of time spent of time spent Not 

in supervision in supervision ascertained N 

Heads of high sections 9 2 112 

Heads of Zow sections 4 7 1 12 

portance, as reflected in statements, such as the following, made by super- 
visors during the interviews 

"I try to understand each girl. I remember I was just one once, and I 
liked to be kind of known by the supervisor. Knowing the girls helps 
with handling the work here you have to know what happens outside 
too to help them inside here at Prudential" (17, p. 10). 

Analysis of the data also showed, as appears in Table 10, that first-line 
supervisors of /n'g/i-production units spent more time on supervision and 
less time on productive work than did /ow-production supervisors. Inter- 
viewers rated section heads on various personality characteristics, with- 
out knowledge of their production records. As appears in Table 11, super- 
visors of /lig/z-production sections were more frequently rated as being 
less arbitrary and less authoritarian than supervisors of /ow-production 
sections. Further analysis of the data indicated that, in general, and in 
so far as the work situation represented by the Prudential Insurance Com- 
pany study is concerned, first-line supervisors of high-production sections 
are significantly more likely to 


"1. receive general rather than close supervision from their superiors; 
"2. like the amount of authority and responsibility they have in their 


"3. spend more time in supervision; 

"4. give general rather than close supervision to their employees; 
"5. be employee-oriented rather than production-oriented" (9, p. 62). 

The Relationship between Supervision 

and the Productivity of Railroad 
Maintenance-of-Way Section Gangs 

Methods employed in the Prudential Insurance Company study have 
been used by the Survey Research Center in an investigation involving 
maintenance-of-way section gangs working in the Pere Marquette Dis- 
trict of the Chesapeake and Ohio Railroad ( 10) . 

The railroad situation resembled that of the insurance company in 
some respects. Both the clerical workers in the latter organization and 
the maintenance-of-way workers on the railroad were generally perform- 
ing low-skill jobs. Both populations tend to be "low in job mobility aspira- 
tions and to be quite satisfied with most aspects of the job situation/' 
However, the two work situations and employee populations differ in 
many important respects, making it possible to investigate, in a very dif- 
ferent setting, some of the factors found to be related to productivity in 
the first study. 

The clerical workers, for example, are not unionized. Most of the rail- 
road workers belong to the Brotherhood of Maintenance of Way Em- 
ployees, although, according to the investigators, "their work situation 
minimizes their contact with the union." The clerical situation is char- 

TABLE 1 1 . Prudential Insurance Company: Characteristics of Supervisors of 

High- and Low-Producing Sections (after Productivity, Supervision and 

Morale in an Office Situation, Survey Research Center, 1950) 

Highly or 

somewhat Not 

Arbitrary reasonable ascertained N 

High producers 
Low producers 







Authoritarian Democratic ascertained N 

High producers 11 1 12 

Low producers 8 4 12 


acterized by white-collar groups, consisting largely of young high-school 
graduates working in large office buildings in a metropolitan area. The 
workers in the railroad situation are all men, mostly over 40 years of age, 
with a median education of from fifth to eighth grade. They perform 
heavy, manual, outdoor labor and live mostly in small villages or towns or 
on farms. Employees in the clerical situation work in close proximity to 
other groups of company employees and to various levels of supervision. 
Contact among the various maintenance-of-way groups in the railroad sit- 
uation is limited, and the contact of members of the gang with upper level 
supervision is restricted to occasional visits from track supervisors. 

The Pere Marquette District of the Chesapeake and Ohio Railroad 
is made up of three divisions, containing a total of about 180 maintenance- 
of-way sections. Each section is headed by a foreman who reports to a 
track supervisor. The latter is in charge of a number of foremen and 
reports, in turn, to a division engineer who heads up the Maintenance- 
of-Way Department for his division. 

The three division engineers and their track supervisors independently 
selected pairs of sections which they judged to be reasonably comparable 
with respect to working conditions, i.e. in terms of terrain, ballast, weight 
of rail, single or multiple trackage, and of other physical factors which 
could produce differences in group performance. In all, 40 pairs of sections 
(of the total of 180 sections) were judged to be very similar with respect 
to such working conditions. 

These sections were then rated on job performance. Working inde- 
pendently, the division engineers and the track supervisors judged which 
section in each pair (under their jurisdiction) was doing the better job 
in terms of "over-all quality and quantity of work performed/' smooth- 
ness of track and alignment of roadbed being given special emphasis as 
criteria of work quality. In the case of 36 of the 40 pairs of sections, the 
division engineer and track supervisor agreed that the same member of 
the pair was superior in performance to the others. The 36 sections con- 
sistently judged to be superior to their counterparts were classified as 
high sections, while the 36 sections consistently judged to be inferior were 
classified as low sections for the purposes of the study. 

The total number of workers employed in the 72 sections was 298, in- 
cluding 156 in the 36 high sections and 142 in the 36 low sections. 4 All 
workers were intensively interviewed by Survey Research Center staff 
members between November, 1948, and January, 1949. Interviews were 
of the free-answer type, and averaged about an hour and one-half in 
length. The 72 foremen of the paired sections were similarly interviewed 

4 Although the sections in each pair are comparable in size, there were small differ- 
ences in numbers of men in the gangs in some cases. 


after the rank-and-file interviewing had been completed. As in the case 
of the study of clerical workers, two major categories of variables (de- 
rived from the interviews) were tested for possible relationship to section 
productivity, viz. 

44 1. The attitudes of the foreman and his relationship to the men in 
his section i.e. how the foreman and men perceive each other, how the 
foreman perceives and fulfills his functions, etc. 

"2. The 'morale 9 of the non-supervisory employees i.e. their atti- 
tudes toward their work and various aspects of their work situations 
(satisfaction with the company, with their fellow workers, with their 
chances for promotions, etc.)" (10, p. 6). 

Analysis of demographic characteristics showed a significant difference 
between foremen of high and low sections in only one respect, viz., more 
foremen of low sections report that they have had some previous job in 
another field while more foremen of high sections have worked only on 
the railroad. On the other hand, findings from interviews with the fore- 
men and rank-and-file employees show differences between the two groups 
of foremen in attitudes and behavior* 

As was also true in the office situation, more of the foremen of the 
high than of low sections report spending a larger proportion of time 
in supervision. Furthermore, a significantly greater proportion of low 
than of high foremen report spending a larger portion of time in straight 
production work. Reasons given by foremen for supervising rather than 
doing production work show that high-section foremen tend to stress 
the effectiveness of supervision in helping their sections do a better 
job, while foremen of low sections are more likely to regard supervision 
as a routine function. In addition, statements obtained from foremen on 
how they spend their working time show, as indicated in Table 12, that 
a larger proportion of foremen of the high sections refer to planning and 
performing special skilled tasks, while a larger proportion of foremen in 
the low sections refer to "doing the same sort of things as the men do." 
Moreover, the foremen's perceptions in this respect are supported by data 
obtained in interviews with rank-and-file employees. As shown in Table 
13, workers in /ng/i-producing sections perceive their supervisors as better 
planners than do men in /ow-producing sections. 

Analysis of the data showed other similarities between the findings of 
the studies conducted in two different settings by the Survey Research 
Center although, in a few respects, the findings of the railroad study failed 
to confirm those of the investigation of clerical workers. An important 
difference in the findings of the two studies is apparent in the area of 

All differences reported in this section, unless otherwise stated, are statistically 
significant at the 5-per-cent level. 



TABLE 12. Chesapeake and Ohio Railroad: Relation to Section Productivity 

of What Foreman Reports Doing on the Job (after Productivity, Supervision 

and Morale among Railroad Workers, Survey Research Center, 1951) 

Supervisory Duties Nonsupervisory Duties 


to men; Same Number 
watching things Keeping of duties 
men Total men do up track Total mentioned * N 

Foremen of 
high sections 42 
Foremen of 
low sections 25 

41 83 8 7 15 98 36 
42 67 15 14 29 96 36 

* Responses total more than 72 because many foremen gave more than 
one answer. 

TABLE 13. Chesapeake and Ohio Railroad: Relation to Section Productivity 
of Men's Perception of Foreman's Planning Ability (after Productivity, Super- 
vision and Morale among Railroad Workers, Survey Research Center, 1951) 

Question: "How good is the foreman at figuring \vorfe out ahead of time?" 

Per cent 




So-so and 
not very good 


Men in 

high sections 
Men in 






low sections 






close versus general supervision. In the clerical situation, there was con- 
siderable evidence that the heads of high sections exercised more general, 
less detailed, supervision over their employees than did the heads of Zow 
sections. Such differences were not found in the railroad situation. Ac- 
cording to the investigators, differences in the structures of the two situa- 
tions may account for the absence of any relationship at all between 
closeness of first-line supervision and productivity, in contrast to the 
negative relationship which appeared in the study of clerical workers. 
In the insurance company, work methods are standardized to such an 
extent that employees get little help of a technical nature from close 
supervision. In the work of the maintenance-of-way section gangs, on 
the other hand, working procedures are less "routinized," and sections 


are small enough to allow the foreman to give each man the benefit of 
his superior technical knowledge. However, the fact that there is no 
positive relationship between productivity and close supervision is inter- 
preted as suggesting that the railroad "foreman's technical contribution 
is not sufficient to outweigh possible detrimental effects of close super- 
vision on worker motivation'' (10, p. 34). 

Data cited above, and other detailed analyses of findings from the rail- 
road study are summarized by the Survey Research Center staff as follows 

"1. High and low foremen do not differ significantly in degree of 
satisfaction with their jobs and other aspects of the work situation. 

"2. Low foremen do not clearly perceive their leadership role. High 
foremen are typically more aware of their position as leader and super- 
visor and are better able to function effectively in this leadership capac- 

44 3. Foremen of high and low sections differ in their attitudes toward 
their men. Foremen of high sections are more positive toward their 
men, take a more personalized approach to them and give more atten- 
tion to problems of their motivation. 

"4. Foremen of high production sections evaluate their sections more 
highly than do foremen of low producing sections" (10, p. 22). 

The Relation between Employee Attitudes 
and Productivity 

The experimental designs of both the study of clerical workers and 
that of railroad workers called for comparisons of the attitudes of workers 
in high and low productivity groups. Analysis of worker attitudes and 
perceptions yielded some of the information concerning differences in 
supervisory behavior discussed in preceding pages of this chapter. In ad- 
dition, a number of items from the employee interviews were combined 
to give a single measure for each of the following attitudinal factors 

l.(a) Pride in work group, defined (only in the study of clerical 
workers) as "the degree of feeling of attachment to and satisfaction 
with the accomplishment of the immediate or secondary work group of 
which employee is a member" (9, p. 39). 

l.(b) Attitudes toward work group, characterized (only in the study 
of railroad workers) as "liking for other men in the section, in their 
feeling that the men stick together well, or in their reactions to the 
good and poor workers in their groups" (10, p. 24). 

2. Intrinsic job satisfaction, characterized as "the degree of satisfac- 
tion obtained by the individual employee from performing those tasks 
which constitute the content of his job" (9, p. 4l). 

3. Company involvement, described as "the degree to which the em- 
ployee derives satisfaction from and identifies with the company in 
which he is employed" (9, p. 43). 

4. Financial and job satisfaction, defined as "the degree of satisfao- 


tion the employee has with his present and expected earnings and with 
the status of his present and expected position in the company" (9, 
p. 44). 

Analyses were made of the interrelationships among these attitudes, 
considered to be significant dimensions of employee morale. The pride 
in work group index, yielding .05 as the highest intercorrelation, proved 
to be quite independent of the remaining measures of attitude. Other 
intercorrelations among attitudinal indices were in the neighborhood of 

Comparison of high- and /ow-productivity groups showed that high- 
and low-producing employees differed consistently only in the evaluation 
of their own work groups. 6 There were statistically significant differences 
in this respect (as measured by the index pride in work group) in the case 
of clerical workers, with the high section indicating a greater degree of 
group pride and loyalty. Similarly, men in the high maintenance-of-way 
sections perceived their work groups as being superior to others (attitudes 
toward work group) although the high-low difference in this case was sig- 
nificant only at the 10-per-cent level. 

In none of the other measures were there consistent differences in at- 
titudes between groups matched for productivity. Furthermore, contrary 
to the findings of the study byJKatz and Hyman (7), referred to in Chap- 
ter 1, the men in the railroad situation who most frequently expressed 
satisfaction with the content of the work (intrinsic job satisfaction) were 
those in the low producing section. A similar, although slight tendency 
appeared in the case of the clerical workers, but differences between high 
and low groups were far short of statistical significance. Of additional 
interest, in the detailed analysis of employee attitudes, is the absence, in 
both groups, of any relation between the employees' liking for the com- 
pany and its policies and practices (company involvement) and produc- 
tivity. Similarly, in the area of satisfaction with job status (financial and 
job satisfaction) there were no differences between employees in high 
and low producing groups. 

The Motivating Potential of Good Supervision 

Findings from the two studies bearing upon employee-attitudes and 
morale are summarized by the Survey Research Center staff in the state- 
ment that "neither study shows as much relationship between employee 
attitudes and group productivity as it shows between supervisory attitudes 

6 In the case of the railroad workers, the possibility was examined that productivity 
differences may reflect differences in education, length of service, etc. No relationship 
was found between such demographic characteristics and production. 


and behavior and group productivity" (10, p. 35). T In reviewing their 
findings, the investigators point out that the data do not supply definitive 
evidence as to "causality" in this situation. Consideration is given to the 
possibility that the personal values and characteristics of the supervisors 
may be functions of the working groups, in the sense that the group which 
produces at a higher rate tends to foster the development of the attitudes 
and behavior traits which characterize the leaders of high production 
sections, and that heads of Zow producing sections become "production- 
centered" because their employees are doing poorly. It is further suggested 
that productivity differentials probably function as reinforcement factors. 
However, according to the staff of the Survey Research Center, "It is 
difficult to see how such personal values and characteristics could be 
produced solely by virtue of having groups who produce at a higher 
rate ... it seems logical that differences in group motivation are related 
to differences in supervisory practice and philosophy" (9, p. 36). 

The Influence of the Supervisor's Role in the Organization: Through- 
out the series of investigations conducted by the Survey Research Center, 
an effort has been made to determine the influence of the relations be- 
tween the first line supervisor and his superiors upon the attitudes and 
behavior of both the first line supervisors themselves and of members of 
their work groups. The studies of office and of railroad workers yielded 
evidence which suggests that supervisors of high production groups are 
more satisfied with the amount of authority they have, and tend to be 
under less pressure than supervisors of Jow production groups. Data from 

T Studies of clerical workers in an insurance company and of railroad maintenance-of- 
way employees have been followed by an extensive investigation conducted in the 
plant of a company manufacturing heavy equipment. Analysis of data is still incom- 
plete at the time of writing. Preliminary, unpublished repoits by Katz & Kahn (8), 
Kahn and Katz (4), Kahn (3), and Neel (14) indicate that findings of the earlier 
investigations concerning the relationship between quality of supervision and produc- 
tivity have been confirmed in this later study. Thus, according to Kahn and Katz, 
"In the three studies of productivity the supervisors with the better production records 
appear to be persons who show in a variety of ways that the individual is important 
to them, that they understand and appreciate him. In brief, the high-producing 
supervisors offer an ego-enhancing relationship. The specific characteristics I have had 
in mind here are the high-producing supervisor's greater time spent in actual super- 
vision, his employee-centeredness, his closeness to or identification with the employ- 
ees, his personal interest in them both on and off the job, his non-punitive behavior 
toward him, and his accessibility for communication" (4). Of additional interest 
in the later study are findings which show that pressure for production when ac- 
companied by consideration of the human relations values involved does not act as 
a deterrent to production. When supervision is a matter of pressure for production 
alone, output suffers. Finally, there are indications that factors such as employee age, 
experience, and job satisfaction, as well as working conditions, interact with super- 
visory practices to influence worker productivity. 


these and other studies, including a large-scale investigation in an elec- 
tric utility company, have indicated also that the attitudes and behavior 
of top management influence the extent to which lower supervisors 
implement the concepts and practices of good human relations on the 
job (13). 

Problems of "organizational climate" raised by such findings are dis- 
cussed in later chapters, especially Chapter 23. Of immediate interest 
for purposes of this chapter is evidence, derived from the electric utility 
company study, that the influence of the supervisor within the organiza- 
tion directly conditions the effect of his behavior on workers' satisfactions 
(15, 16). Analysis of data from earlier phases of this study showed that 
employees in small work groups (10 employees or fewer) thought highly 
of a supervisor who "sides with them" in case of conflict with higher 
supervision. By contrast, in large white-collar groups, "employees were 
significantly less satisfied with such a supervisor; they preferred a super- 
visor who sided with management" (15, p. 52). 

These and other findings led to a further investigation of the factors 
which lead employees to accept a supervisor, in the sense of following the 
leader's suggestions, expressing satisfaction with his conduct, etc. For 
purposes of this study it was postulated that the "supervisor will be 
'accepted' by his work group if -iiis behavior helps them to achieve their 
goals" (15, p. 53). The important element in this situation is the extent 
of the supervisor's influence within his department, since the attempts of 
a supervisor to help employees to reach their goals will succeed if he has 
influence in his department; they will fail if he lacks such influence. 

Supervisors were classified into "high influence" and "low influence" 
groups and comparisons made between measures of their behavior and 
employee attitudes. An examination of correlational indices showed that 
under high-influence supervisors, "siding with employees" and "social 
closeness" were accompanied by some rise in employee satisfaction in 
19 out of 28 comparisons, and that 7 of the positive correlations were 
statistically significant at the 5 per cent level, a result which is 1 1 times 
better than chance. Under low-influence supervisors, such supervisory 
conduct was accompanied "by a rise in employee satisfaction only 8 
times out of 28; a loss in satisfaction (negative effect) is the more com- 
mon result" (16, p. 216). The correlation values were modest in size, 
but, according to Pelz, positive results under influential supervisors were 
sufficiently uniform that the total set may be regarded as highly trust- 
worthy" (16, p. 216), in terms, at least, of statistical significance. "It 
seems fairly clear/' it is added, "that a supervisor's influence or power 
within the department does 'condition' the way his supervisory behavior 
affects employee attitudes. It is plausible to conclude that the supervisory 


behaviors of 'siding with employees' and 'social closeness to employees' 
mil tend to raise employee satisfaction only if the supervisor has enough 
influence to make these behaviors pay off in terms of actual benefits for 
employees" (16, p. 216). Furthermore, consideration of the situation in 
the electric utility company showed that "in this company's small groups 
the supervisors who sided with employees also tended to be the ones who 
had high influence and could get results when they attempted to do 
something. On the other hand, in this company's large work groups the 
supervisors who took the employees' side were generally the ones without 
influence; they could not follow up their helpful attempts with concrete 
gains for employees; as a result, employees were less satisfied than if 
their supervisors had maintained a neutral position" (16, p. 216). 

In contrast to the conclusions reported in earlier sections of this chap- 
ter, those on the effects of the influence of the supervisor are derived from 
research in a single company. For this, and other reasons, caution must 
be exercised in terms of generalizations to industry at large. Furthermore, 
the evaluation of these findings from an administrative viewpoint must 
include a consideration of how far management can go in satisfying work- 
ers where the satisfaction of employee goals conflicts with the satisfaction 
of other goals which must be achieved by the business enterprise. This 
aspect of the situation the general problem of conflicting personal and 
organizational goals is discussed in later chapters, especially Chapter 
23. In the meantime, although much remains to be done in the further 
analysis of a complex situation, the studies described above clearly 
demonstrate that the attitudes and behavior of the first-line supervisor 
are important factors in determining the productivity of a work group. 
They also provide important clues to the characteristics of the supervisor 
capable of making a definite contribution to the attainment of produc- 
tivity goals. In particular, it is clear from these studies that the high pro- 
ducing supervisor looks upon the problem of productivity as one of motiva- 
tion, seeing his role as being primarily that of "motivating workers to 
achieve a goal, of creating conditions under which the goal can be 
achieved" (10, p. xi). 

In Summary: It is evident that the achievement of industrial objectives 
discussed in Chapter 1 requires close attention to the quality of super- 
vision in industry. This need is likewise apparent from studies discussed 
in Chapter 7, and from the findings of employee attitude surveys de- 
scribed in Part III of this volume. Specifically, it appears that the super- 
visor who sees the problem of productivity exclusively in the technical 
terms of work methods and standards is less likely to motivate workers 
to increased production than one who sees the problem in terms of work- 
ers' status, characteristics, needs, and aspirations. The development of 


"employee-oriented" supervisors, adequately equipped for dealing with 
interpersonal relations on the job, therefore represents one promising ap- 
proach to the solution of problems of motivation and morale in industry. 


1. H. Feldman, Problems in Labor Relations. The Macmfllan Co., 1937. 

2. Human Relations Program of the Survey Research Center First Three Years of 
Development. University of Michigan: Institute for Social Research, 1950. 

3. R. L. Kahn, The Importance of Human Relations Research for Industrial Produc- 
tivity (Mimeographed report of the Survey Research Center, University of 
Michigan) 1951. 

4. R. L. Kahn and D. Katz, Some Relationships Between Organizational Character- 
istics and Productivity (Mimeographed report of the Survey Research Center, 
University of Michigan) 1952. 

5. D. Katz, "Morale and motivation in industry" in Current Trends in Industrial 
Psychology. University of Pittsburgh Press, 1949, 145-71. 

6. D. Katz, "The Role of quantitative research in group organization and group 
functioning" in Research on Human Relations in Administration. University of 
Michigan, Institute for Social Research, March 11, 1949. 

7. D. Katz and H. Hyman, "Morale in war industry" in T. Newcomb and E. Hartley 
(Eds.), Readings in Social Psychology. Henry Holt & Co., 1947, 437-47. 

8. D. Katz and R. L. Kahn, Some Recent Findings in Human Relations Research 
(Mimeographed report of the Survey Research Center, University of Michigan) 

9. D. Katz, N. Maccoby, and N. C. Morse, Productivity, Supervision and Morale 
in an Office Situation. University of Michigan: Survey Research Center, 1950. 

10. D. Katz, N. Maccoby, G. Gurin, and L. G. Floor, Productivity, Supervision and 
Morale among Railroad Workers. University of Michigan: Survey Research Cen- 
ter, 1951. 

1 1 . R. Likert, A Program of Research on the Fundamental Problems of Organizing 
Human Behavior. University of Michigan: Survey Research Center, 1947. 

12. N. Maccoby, "The Relationship of supervisory behavior and attitudes to group 
productivity in two widely different industrial settings" in APA Papers (Presented 
by Staff Members of the Institute for Social Research). University of Michigan: 
Survey Research Center, September, 1949. 

13. F. C. Mann, "Changing superior-subordinate relationships/' /. Soc. Issues, 1951, 
7, 3, 56-63. 

14. R. G. Neel, Factors Related to Productivity (Unpublished report, Survey Re- 
search Center, University of Michigan) 1952. 

15. D. C. Pelz, "Leadership within a hierarchical organization/' /. Soc. Issues, 1951, 

16. D. C. Pelz, "Influence: a key to effective leadership in the first-line supervisor," 
Personnel, 1952, 3, 209-17. 

17. Productivity, Supervision and Employee Morale (Human Relations, Series i, 
Report 1). University of Michigan: Survey Research Center, 1948. 

18. E. D. Smith, "The minor executive and mental hygiene" in H. B. Elkind (Ed.), 
Preventive Management. B. C. Forbes & Sons Publishing Co., Inc., 1931, 197- 

Employee Participation in 

IN THE operation of a business enterprise, executives and 
supervisory personnel are faced daily with the need for making decisions. 
Many, if not all such decisions, either directly affect workers or are of 
interest to them because of the potential effects upon their daily activi- 
ties and ultimate welfare. Workers, and also subordinate supervisory 
personnel, "may have a strong desire, particularly in a nation with deeply- 
ingrained democratic traditions, to participate in the determination of 
matters affecting them" 1 (23, p. 410). As a result, it is important to 
consider both the influence which employee participation in decision- 
making may have upon motivation and morale and the form which such 
participation might best assume (22a, 23). 

Investigations bearing upon productivity, supervision, and morale by 
the Survey Research Center, described in Chapter 8, led Katz to conclude 
that workers "do better when some degree of decision-making about their 
jobs is possible than when all decisions are made for them" (11). An 
elaboration of the underlying psychological processes is suggested by 
Maccoby who, following a review of findings from the studies, concludes 
that "Employees may be motivated by repeated external stimulation by 
the supervisor who is constantly instructing his employees on what highly 
specific task to do next . . . On the other hand, if a situation can be 
established, whereby the individual can for the most part furnish his own 
motivation, where stimulation comes from the perception of a larger and 
more inclusive task, the goals of which are clear to him and accepted by 
him, 2 then a higher level of performance and a higher level of morale may 
be achieved ( 16, p. 5) . 

1 Not italicized in the original. 
1 Not italicized in the original. 



This conclusion is in accord with that derived from earlier studies, con- 
ducted by Lewin and his coworkers in the area of group dynamics, dis- 
cussed in Chapter 6. A major outcome of such research is the conclusion 
that employee participation in decision-making, in a democratic atmos- 
phere created by "permissive" leadership, facilitates the development of 
"internalized" motivation, and serves to raise the levels of the employee 
production and morale ( 1, 2, 7) . 

Studies in the Norwood Manufacturing 
Company Plant 

Evidence in support of this position is found in a series of studies con- 
ducted in the Harwood Manufacturing Company plant by a number of 
psychologists, including Marrow (president of the company), Lewin, 
Bavelas, French, and Coch. Production workers in this garment-making 
plant were paid a piecework rate based on standards established through 
the use of time and motion study (9). The standard production rate for 
each job was set at 60 units per hour representing, e.g., 120 dozen button- 
holes or 40 complicated fitting operations. Basic wage rates of 72 cents 
per hour were in line with area scales for the industry, and workers ex- 
ceeding the standard were paid 100 per cent of the labor savings effected. 


Among the labor problems in the plant was resistance on the part of 
production workers to necessary changes in jobs and work methods. Such 
resistance was evidenced in a number of ways. For example, for the sim- 
plest type of job in the plant the average learning time for beginners 
was five weeks. When transferred to the same job, experienced operators 
required an average of eight weeks to reach standard production. More- 
over, employees involved in job transfers and related changes in job 
methods exhibited a general pattern of low morale, characterized by "ex- 
pressions of resentment and aggression against management . . . with 
evidences of frustration, loss of hope of ever regaining the former level 
of production and status, feelings of failure, and a very low level of 
aspiration" (21, p. 346). 

According to Marrow, it could not be contended that skill habits de- 
veloped on the original job were interfering with relearning on the new 
job, since time-and-motion studies showed yery few false moves after 
the first week of change, and transferred operators rarely showed a desire 
to go back to the old way of work. From such evidence it was deduced 
that retroactive inhibition [sic] could not account for the slow rate of 
learning the new job methods (20). Furthermore, efforts to solve the 


problem by means of monetary allowances, by layoffs, etc., were of no 
avail. On the basis of these and other considerations, it was concluded 
that the situation represented a deep-seated motivational problem re- 
flecting resistance to change which, it is contended, appears with par- 
ticular strength in groups with a "we" or "ingroup" feeling antagonistic 
to management and its objectives, and which may find overt expression 
in restriction of output (5, 28). 

On the basis of such general theory, an experiment was devised to 
examine the effect of group participation in setting production goals 
upon overcoming resistance to change and in arousing attitudes and moti- 
vation conducive to meeting production standards. Four groups of work- 
ers involved in changes in job methods and rates were selected for the 
experiment. The groups were similar in terms of group efficiency ratings 
prior to change. The degree of change to be made and the amount of 
"ingroup feeling" observed were the same for all groups. 

In the case of one group, constituting a control group, the usual factory 
routine was followed in making the job changes. The production de- 
partment modified the job, and a new piece rate was set. A group meeting 
was held in which the control group was told that the change was 
necessary because of competitive conditions, and that a new piece rate 
had been set. The new piece rate was explained by the time-study man, 
questions were answered, and the meeting was dismissed. 

"Democratic participation" methods were used in arriving at changes 
in the case of the remaining three groups, constituting the experimental 
groups. Before any changes took place, a group meeting was held with 
the operators to be changed. The group was told that changes in the job 
were required by the sales department. General agreement was reached 
that savings could be effected by removing some "frills" and "fancy" 
work from the garments involved without affecting the opportunity to 
achieve a high efficiency rating. Furthermore, provisions were made for 
the groups to participate in planning the details of the change. Group 1 
of the experimental groups was represented in the planning by selected 
"special" operators. All members of experiment Groups 2 and 3 partici- 
pated in the planning. 

As is evident from Figure 13, production of the control group dropped 
immediately after the job changes were made and training completed. 
Furthermore, this group showed practically no improvement in efficiency 
by the end of an experimental period of 40 days. Resistance developed 
almost immediately after the changes were introduced. There were 
marked instances of aggression against management, including conflict 
with the methods engineer, deliberate restriction of production, and lack 


CM ** 

a CL 

QC Ql 03 

s - ^ 

a o: * ^ 

XX (3 2" 

ui u .0 -- 1 

PH ^ 



O ~ 


^ O 


z cr 
o x 


S "5 

C3 "^ 

o ^* 

a. 3 









CxO o 

;* P^ 





of co-operation with the supervisor. Within 15 days after the change, 
9 per cent of the operators quit, and a total of 17 per cent quit within 
40 days. Grievances were filed about the piece rate, although the latter, 
upon checking, was actually found to be a little "loose." 

Marked recovery to and beyond prechange levels of efficiency was 
achieved in the case of all three experimental groups. As shown in Figure 
13, Groups 2 and 3 reached their prechange level of production the 
second day after change, and by the end of the experiment they had 
actually surpassed their prechange output level by about 14 per cent. 
They worked co-operatively with their supervisors; there were no signs 
of aggression, and there were no quits, within the first 40 days following 
the completion of training on the new job methods. Group 1 also ex- 
ceeded its prechange level of production by the end of the 40-day ex- 
perimental period, although it required more time to recover to the pre- 
change production level, possibly, according to Marrow, because of an 
unavoidable operational problem. No quits were recorded for this group. 
One act of aggression was observed, and this was neither prolonged nor 


Additional evidence that participation in decision-making can help 
to increase productivity is derived from another experiment conducted 
by Bavelas, in the plant of the Harwood Manufacturing Corporation. 
In this study, as reported by French (6), meetings were conducted with 
small groups of sewing-machine operators, including 4 to 12 workers. In 
the course of such meetings, the experimenter (trained in psychology 
and also experienced in dealing with labor) talked about the greater 
ease of working together as a team; discussed individual production 
levels with the group; questioned its members as to the level of produc- 
tion which might be obtained if they worked as a team, and asked if 
they would like to set a team goal for higher production. In addition, 
an effort was made to obtain a group decision as to the level of the new 
production goal and as to the time in which it should be achieved by the 

Discussions, according to French, were marked by a "friendly style 
of democratic leadership/' and the experimenter "was careful to leave 
all decisions up to the group, namely, whether they wanted to set a goal, 
and if so, at what level and in what time." In some of the groups, desig- 
nated as control groups, no such decisions were made. In most of the 
groups, the members decided to set a team goal for increased produc- 
tion and specified the time in which the goal would be achieved. In 
the case of such groups, designated as experimental groups, the leader 


usually arranged for further meetings to check on progress, during which 
he provided them with graphs showing changes in productivity for the 

The experiment covered a period of four months. During this time, 
the average efficiency rating for the total plant varied between 52 and 
58 units per hour. Both the control and experimental groups were made 
up of operators who were superior in production, with efficiency ratings 
(for the groups) ranging from 64 to 68 units per hour. Each control 
group was matched with an experimental group with respect to type of 
job, type of "social setting" and supervision. Analysis of production 
records showed that in the case of the control groups the level of produc- 
tion remained relatively constant over the four-month period. In the 
case of the experimental groups there was, in contrast, an increase in 
production, averaging 18 per cent, which was consistently maintained 
for the period of over two months subsequent to the group discussions 
of production goals. This increase, it is reported, was obtained without 
any changes in job methods or other physical conditions of work, and 
can be ascribed "to the leadership and the social factors involved in group 
decisions/' According to French, the results of this and of subsequent 
experiments, involving both group decision and the use of "pacing cards/' 
justify the conclusion that "the application of skillful democratic tech- 
niques of leadership in industrial settings can result in extremely marked 
increases in group productivity 'which 'will persist over long periods of 
time" 9 (6, p. 87). 


Experiments of the types described above, conducted in the factory 
setting, are of extreme interest from the viewpoint of motivational theory. 
They also have great practical import, particularly in view of the current 
drive for employee participation in what have been traditionally viewed 
as managerial functions in industry. 4 This underlines the necessity, basic 
to any research, of making certain that the influence of all variables 
which might account for the observed changes be taken into considera- 
tion both in the design of the experiment and in the interpretation of 
findings. The experimental designs of the studies conducted in the Har- 
wood Manufacturing Company plant appear to be defective in these 
respects. The design for neither study, as reported by the investigators, 
eliminated the possibility that variables other than participation in de- 
cision-making played a role in accounting for the production differences 
between the control and the experimental groups. 

* Not italicized in the original. 

* See Chapter 23. 


Of special importance, in the first of these studies, is an apparent dif- 
ference in the character of the training received by operators in the two 
groups. While none of the available reports on this study are particularly 
clear with respect to this issue, it quite definitely appears that more and 
better training was received by members of the experimental groups than 
by members of the control groups. In the second study, as reported by 
French, it seems quite evident that each of the experimental groups was 
supplied with considerable knowledge of results, in the form of graphs 
showing changes in the productivity of the group. There is no indica- 
tion that similar information was supplied to the control groups. As 
shown in Chapter 7, knowledge of results can in itself lead to an increase 
in production, There is considerable evidence, from experiments in the 
training of industrial workers, that better-trained groups perform better 
than do those who are not given the advantage of systematic training 
(8, 24). Failure to control adequately the possible effects of difference 
in training, in knowledge of results (and possibly of other variables) 
makes it impossible to accept the view that the increased output by the 
experimental groups in the Harwood plant studies was necessarily due 
solely to employee participation in decision-making. 5 Nevertheless, the 
general nature of the experiments and of findings from these and other 
studies still support the view that employee participation in decision- 
making can contribute to increased employee productivity. 

A Case Study from the Tavistock Institute 
of Human Relations 

A case study from investigations conducted by the Tavistock Institute 
of Human Relations, in Great Britain, furnishes another illustration of 
research on worker participation in decision-making conducted by pro- 
ponents of Lewin's "field-theoretical" approach. 8 In the Service Depart- 
ment of the Glacier Metal Company there was a chronic condition of low 
morale, caused by the piecework system of wage payment in use, under 
which some unskilled workers were earning more than craftsmen. With 
the consent of both workers and management, a research team under- 
took the task of investigating the morale problem in this department, 

8 It is of interest to note that a similar defect in experimental design appears also 
in an earlier experiment by Bavelas (la) on democratic vs, authoritarian leadership, 
which anticipated the program of group-dynamics research conducted in the plant 
of the Harwood Manufacturing Company. As Preston has pointed out, the latter 
experiment did not exclude "the possibility that the [observed] differences in efficiency 
and morale are due to the differences between lack of training and effective training 
[of leaders], rather than to the difference between authoritarian and democratic 
techniques" (2 la, p. 27). 

See Chapter 6. 


and of dealing also with problems arising from a proposed changeover 
from a piece rate to an hourly-wage method of payment. 

There were mixed feelings about the proposed change, arising in part 
from the fact that some of the 40 workers on piece-rate pay would re- 
ceive more pay, and others less than under the existing wage system. 
Many of the workers were suspicious of management's motives. With 
the aid of the research team a Wages Committee including in its mem- 
bership an existing Shop Committee, the division manager, the shop 
superintendent, and the shop accountant was set up to study the pos- 
sibilities of this change. To help allay suspicion and to clarify attitudes, 
group discussions were inaugurated at the shop-floor level. For this pur- 
pose, the 40 piece-rate workers were divided into five groups, each com- 
posed of eight workers with a member of the Shop Committee delegated 
to each. 

After approximately three months of discussion and group meetings 
the first secret ballot was taken on the proposed changeover. The 40 
workers voted unanimously to have their representatives carry out further 
discussion in the Wages Committee. During these meetings it became 
apparent, according to Jaques, that in order to arrive at a satisfactory 
solution to the wages question, problems affecting the morale of the 
shop would have to be dealt with, including such matters as "1 . A guaran- 
tee from the shop that production would not suffer; 2. A guarantee from 
management to the shop re benefits from increasing productivity; 3, The 
establishment of some mechanism for making possible increased partici- 
pation of the total shop in the making of departmental policy" (10, 
p. 243). 

On a second ballot, taken after another month of discussion, 28 of 
the 40 workers voted in favor of the proposed changeover from a piece 
rate to an hourly-wage method of payment. On the same ballot was a 
statement that open questions as to the effects of productivity changes, 
savings in overhead, etc., upon the new method of wage payment would 
be left to "the findings of a Shop Council (to be set up as representative 
of all London Service Station personnel)/' According to Jaques, the 
significant outcome of the discussion sessions arranged by the research 
team was this decision to establish a "Shop Council which gave promise 
of being a mechanism through which members of the department could 
take part in policy making, and in so doing mediate in inter-group rela- 
tions in the department" (10, p. 249). 

Since the issue of the Shop Council was not separated on the ballot 
from the change in the wage system, it is difficult to know to what 
extent the voting was influenced independently by (1) the desire for 


participation in decision-making, and (2) the desire for a change in wages. 
This is a particularly crucial question in view of the fact that the "yes" 
votes may well have been from men who stood to profit by increased 
wages from the changeover from the piece rate to the hourly-wage 
method of payment. 

An Investigation in a Telephone Company 

Further evidence with respect to the significance of participation in 
decision-making is found in an investigation of turnover and morale 
among young women employees of the Michigan Bell Telephone Com- 
pany (26). Subjects of the investigation were all of the operators and 
service representatives, approximately 600 in number, hired in the cities 
of Lansing and Grand Rapids, and 97 operators hired in Pontiac during 
the period from January, 1945, to February, 1948. 7 These were divided 
into two groups including (a) those still in service at the time of the 
study, designated as on-force, and (b) those who had been separated 
from the company, designated as losses. 

Questionnaires, designed to measure attitudes towards the girls' jobs, 
were administered, in groups and on company time, to practically all of 
the on-force operators in Lansing, Grand Rapids and Pontiac and to the 
on-force service representatives in Lansing and Grand Rapids. The same 
questionnaire, with necessary changes in the tense of verbs, was admin- 
istered individually, in their homes, to as many of the Lansing and Grand 
Rapids operators and service representatives who had left the Company 
as could be located. In addition, "depth" interviews were conducted 
with service representatives and operator losses.* 

Comparisons between on-force employees and those who had left the 
Company (i.e. losses) showed that the two groups were not significantly 
differentiated by biographical data, by employment test scores, or in 
terms of "neurotic tendency" as measured by scores on "some personality 
test items" included in the questionnaire. Likewise, the groups were not 
differentiated in terms of attitudes frequently considered to be important 
in accounting for turnover. For example, there were no consistently 
significant differences between the two groups in attitudes toward wages. 
In fact, there were indications of a trend for young women who had left 
to "feel better about their wages" than those who remained in the 

T Exclusive of those on maternity leave in March 1948. 

8 Of these girls, 56, or 47 per cent, of the Lansing operators losses were located, 
as were 47, or 68 per cent, of the Grand Rapids operators losses and 25, or 71 per 
cent, of the Grand Rapids service representative losses. Only 8, or 19 per cent, of 
Lansing service representative losses could be traced, so this group was dropped from 
the statistical treatment of findings. 


service of the company. For the most part, those who had left were not 
any more critical of the policies of the company than those who stayed 
on the job. 

Replies showed a more favorable attitude toward supervision on the 
part of on-force personnel to an extent which, according to the investi- 
gator, is "sufficient to deserve some attention/' Of particular significance 
and interest are findings which show that those who stayed on the job 
tended to feel that (a) they had a chance to make decisions on the job 
and that (b) they were making an important contribution to the success 
of the company. 

TABLE 14. Michigan Bell Telephone Company: Employee Attitudes toward 
Participation in Decision-Making (after Wickert) 

Answers to Question: "Do you (or did you) have a chance to make decisions 
on your job at Bell?" 

"Yes, very often" 


or "sometimes" 

or "never" 

per cent 

per cent 



Lansing * 

On Force 78 22 88 

Losses 42 58 43 

Grand Rapids * 

On Force 74 26 125 

Losses 45 55 40 


On Force 73 27 97 


Service Representatives 


On Force 81 19 26 

Losses t 

Grand Rapids * 

On Force 86 14 36 

Losses 38 < 62 26 


* Differences between percentages significant beyond the one-per-cent level 
of confidence. 

t Not reported insufficient number of cases. 


As is apparent from Table 14, in the case of all groups for which com- 
parisons could be made, there is a sizable and statistically significant dif- 
ference between on-force personnel and losses with respect to the fre- 
quency with which those who stayed on the job answered "yes, very 
often' 9 or "sometimes" (rather than "seldom" or "never") to the ques- 
tion "Do you (or did you) have a chance to make decisions on your job 
at Bell?" Similar differences were found in response to a question dealing 
with the feeling of contribution to the success of the company. Accord- 
ing to Wickert, who conducted the study, the findings show that "the 
one, clear-cut marked difference was that the girls who stayed [as con- 
trasted with those who left the job] seemed more ego-involved in their 
work . . . Here" he adds, "is a bit of empirical evidence in favor of in- 
dustrial democracy" 9 (26, p. 197). 

An Experiment in Group Participation 
in Decision-Making by Supervisory Personnel 

An experiment by Levine and Butler (12) shows that group discus- 
sions (representing one medium for participation in decision-making) 
can be useful at the supervisory level. 

The subjects of this investigation were 29 supervisors of 395 workers 
in a manufacturing plant who were paid an hourly rate. Every six months 
the workers were rated by their supervisor on five factors, viz., accuracy, 
effective use of working time, output, application of job knowledge, and 
co-operation. The ratings were then used in determining each worker's 
hourly rate. 

An analysis showed that supervisors' ratings were influenced by a "halo 
effect" in that performance on the job by the individual worker was not 
being rated alone, but that the grade of the job (whether high or low) 
was a determining factor in the ratings given. As a result, workers in the 
higher-grade jobs were paid the highest of their respective wage rates, 
while those in the lower-grade jobs consistently received the lowest of 
their respective wage rates. 

The object of the experiment was to compare the effectiveness of 
group decision versus formal lectures in inducing the supervisors to over- 
come their biased performance ratings. Supervisors were randomly divided 
into three groups of 9, 9, and 11. Group A served as a control group in 
that supervisors in this group continued making ratings without any 
special instructions. Group B approached the problem from a discussion 
viewpoint and Group C was given a formal lecture. Only Group B (the 
discussion group) showed any significant change in ratings. Both the con- 
trol group and the lecture group continued to make biased ratings, while 

9 Not italicized in the original. 


the ratings of Group B underwent modification to the point where simi- 
lar distributions of ratings were achieved for workers in high- and low- 
grade jobs. The authors view their findings as confirmation of the earlier 
work of Lewin in demonstrating the greater effectiveness of group de- 
cision over a lecture method of training. 

The Influence of "Permissive" vs. 

''Restrictive 7 ' Leadership upon Job 

Satisfaction, Productivity, and Morale 10 

Experiments conducted at the National Training Laboratory for Group 
Development, by the Research Center for Group Dynamics, with the 
support of the Office of Naval Research, have suggested the possibility 
of applying research techniques for the study of "autocratic" and 
"democratic" atmospheres to problems of intellectual productivity and 
of team work at management levels (4, 6, 23). Of particular interest, 
in this connection, is an investigation of interpersonal relations in a naval 
research laboratory (25), conducted by the Human Relations Research 
Group, Institute of Industrial Relations, University of California (Los 
Angeles), under a grant from the Office of Naval Research. 

Two divisions, designated as Division A and Division B, were studied. 
Employees in both divisions included physicists, engineers, and scientific 
aides, with supporting clerical personnel, numbering 28 in Division A, 
and 38 in Division B. The two divisions differed in the style of leadership 
under which they operated and in the "atmosphere" which appeared 
to prevail in each. The head of Division A was a brilliant young scientist 
who directed the division along restrictive lines. Division B was under 
the leadership of an older man, a fatherly type, who directed his division 
along permissive lines. The manner in which these two division leaders 
operated (i.e. the behavior of each division head) was "operationally 
observed through a collection of specific incidents which characterized 
his activities" derived from interviews and comments included in socio- 
metric questionnaires. 

Each subject from the two divisions was asked, as part of a larger 
investigation, to indicate on a questionnaire his level of job satisfaction, 
and his perception of the level of productivity and of morale of his own 
work group, of his division and of the laboratory as a whole. Job satis- 
faction referred to the individual's personal satisfaction with his position, 
as evaluated by himself on a five-point scale, from "very high" (scale 

10 It has seemed desirable to discuss this investigation in the context of participa- 
tion in decision-making (as reflected in the concepts of restrictive and permissive 
leadership), although the findings are also significant in terms of interrelations be- 
tween supervision, on the one hand, and job satisfaction, productivity, and morale 
on the other, as discussed in Chapter 8. 


value 1) to "very low" (scale value 5) . With respect to productivity, each 
individual was requested to evaluate, on a similar five-point scale, the 
level of production of his own work group, of the division and of the 
laboratory as a whole. (It should be noted that this rating reflects 
productivity as perceived by members of the group and, as will appear 
later, is not necessarily an index of actual production or of productivity 
level as evaluated by superiors.) Following the pattern established in an 
earlier study by Bernberg (3), which showed that "the collective rating of 
a group of employees concerning their own morale was the most mean- 
ingful single variable related to other criteria of morale," each subject 
was asked to rate the level of morale of his group, of his division and of 
the laboratory on the five-point scale. While the specific interpretation 
of the word ''morale" was left to each individual, the related question- 
naire item "implied a group concept as distinguished from individual 
job satisfaction." As indicated in Figure 14, analysis of the findings 
showed that 

1. A considerably higher proportion of the members of Division B, 
headed by a permissive leader, considered themselves to be satisfied with 
their jobs. 

2. The majority of individuals in both divisions rated the productivity 
of their work groups as above average. However, a larger proportion of the 
members of Division B rated the productivity of their division and the 
laboratory as either 'Very high" or "high" than did members of Division 
A, headed by a restrictive leader. 

3. A strikingly higher proportion of the members of Division B con- 
sidered the level of morale of their work group, of their division, and 
of the laboratory to be "very high" or "high" than was the case in 
Division A. 

The data so far reported reflects the perceptions of members of the 
two divisions with respect to productivity, job satisfaction, and morale. 
In addition, interviews were held with five superiors and two staff mem- 
bers, all familiar with the objectives and performances of both divisions, 
in an effort to arrive at a more objective, critical comparison of the two 
units. Data obtained in this way included evaluations of the productivity 
of both divisions. These proved to be of special interest, since, in terms 
of the standard set by the superiors, Division A was judged to be per- 
forming an adequate job in spite of the restrictive leadership and, as a 
matter of fact, a better job than the group gave itself credit for. More- 
over, the superiors rated the performance of Division B lower than did 
members of the division. 

Interviews with the seven key individuals and with division supervisory 
personnel provided a partial explanation of these discrepancies. In the 
operation of the laboratory, each division head exercised considerable in- 



fluencc in determining objectives for his division. It appeared, however, 
that the objectives which the superiors assumed for Division A were 
more or less identical with those set by the head of this division, who 
was able to lead his group in that direction in spite of low morale and 


Above- Job Perceived Productivity Perceived Morafe, Above- 

Average Satisfaction WorkGroup Division Laboratory WorkGroup Division laboratory Average 


81.6 81.6 




































7.9 *6^ 








"a iSi 


^ n.9 












Figure 14. Bar Chart Comparing Percentage Above- Average [Very High ( 1 ) 
and High (2)] Ratings with Percentage Below- Average [Low (4) and Very 
Low (5)] Ratings for Job Satisfaction, Perceived Productivity, and Perceived 
Morale (Divisions A and B) * 

job satisfaction. On the other hand, the objectives set by the head of 
Division B did not correspond too closely with those assumed by his 
superiors. As a result, he "utilized the services oj a high morale group and 
of satisfied people in the performance of tasks which his superiors did not 
consider of highest importance to the Laboratory' 9 X1 (25, p. 6) . 


Detailed analysis of findings in the nav^l research laboratory study 
supports the observation, reported in Survey Research Center investiga- 
tions described in Chapter 8, concerning the lack of relationship be- 

11 Not italicized in the original. 

* By permission, from I. R. Wcschlcr, M. Kahanc, and R. Tannenbaum, "Job 
satisfaction, productivity and morale: a case study/' Occup. Psychol., 1952, 26, 1-14. 


tween productivity, on the one hand, and job satisfaction and morale, on 
the other. Such findings, combined with those discussed above, focus 
attention upon an important practical problem which needs careful 
thought in considering the place of employee participation in decision- 
making in the industrial plant. In particular, the difference between 
the group perception of the situation and that of the superiors or "man- 
agers" of the laboratory suggests that conflicts can (and do) arise be- 
tween organizational goals and personal goals. 

As is suggested by those responsible for the naval research laboratory 
study, "a unit which has been striving to set a community-chest record 
may get its gratification through this type of success rather than through 
a spectacular showing in productivity" (25, p. 7). Similarly, individuals 
in a factory group may gain considerable satisfaction from participation 
in decision-making, and the level of group morale may rise to consider- 
able heights. At the same time, productivity, considered an equally im- 
portant goal by a management faced with the need of meeting the pay- 
roll and other business expenses, may remain unaffected. 

In closer detail, consultation with a work group concerning the details 
of a job evaluation or incentive plan may result in decisions acceptable 
to the group which, however, run counter to the best judgment of the 
job-evaluation expert, the time-and-motion-study man, or of top super- 
vision. As Whyte has suggested, "groupthink" can lead to mediocre as 
well as to superior solutions of organizational problems and of the broader 
problem of society (27). However, there are distinct indications that 
participation in decision-making can yield adequate solutions and that 
such participation tends to bring about a coincidence of formal (i.e. or- 
ganizational) and personal goals (23, p. 414). Nevertheless, further study 
is needed to determine more exactly the "dynamic psychological con- 
ditions" under which "the incorporation of formal goals within the 
scope of the personal goals" occurs, and also as to the optimal limits of 
participation at individual and group levels (23). There are problems in 
the situation which must be faced and resolved, not by glib use of the 
concept of "democratic leadership," but by continued research which can 
yield greater insight into both the role and the limits of employee par- 
ticipation in decision-making, within a frame of reference which rec- 
ognizes the necessity of increased industrial productivity, as well as of 
individual employee satisfaction and group morale, in achieving the 
objectives of a democratic society. 


Group decisions of the types involved in experiments described in this 
chapter are possible only in an organization which permits employee 


participation in making decisions (1). According to the exponents of 
the group-dynamics approach, the benefits of group decisions are there- 
fore to be obtained only within the framework of a democratic group 
structure in industry which, it is claimed, is the key to reduced labor 
conflict as well as to increased production. In this connection, there 
arises immediately a series of questions concerning the way in which 
such a democratic group structure can best be achieved* (13, 14, 18, 19) . 
Involved also is the question as to whether employee participation should 
(and can) be limited to decisions on production problems or should in- 
clude a large variety of plant problems of concern to employees as well 
as to management (15, 17). This question merges into broader issues 
arising in formal systems of joint consultation through the agency of 
labor-management or union-management committees. Such issues, in- 
cluding that of codetermination (22), are discussed in Chapter 23, which 
includes a consideration of the role of formal systems of participation in 
the stimulation of motivation and morale in industry. 


L A. Bavelas, "Some problems of organizational change," /. Soc. Issues, 1948, 4, 

la. A. Bavelas, "Morale and training of leaders" in G. Watson (Ed.), Civilian 
Morale. Houghton Mifflin Co., 1942, 143-65. 

2. A. Bavelas and K. Lewin, "Training in democratic leadership/' /. Abnor. Soc. 
Psychol., 1942,37, 115-119. 

3. R. E. Bernberg, An Objective Analysis of Some of the Socio-Psychological Factors 
in Industrial Morale. (Unpublished Ph.D. Dissertation.) University of California, 

4. L. Bradford and }. R. P. French, Jr., "Introduction/' in "The Dynamics of the 
Discussion Group," /. Soc. Issues, (Spring Issue) 1948, 4, 2, 2-9. 

5. L. Coch and J. R. P. French, Jr., "Overcoming resistance to change/' Hum. 
Relat., 1948, 1, 512-32. 

6. J. R. P. French, Jr., "Field experiments; changing group productivity" in J. G. 
Miller (Ed.), Experiments in Social Process. McGraw-Hill Book Co., 1950, 83- 

7. J. R. P. French, Jr. and A. Zander, "The group dynamics approach" in A. Korn- 
hauser (Ed.), Psychology of Labor Management Relationships. Champaign, 111.: 
Industrial Relations Research Assn., 1949, 71-80. 

8. E. E. Ghiselli and C. W. Brown, Personnel and Industrial Psychology. McGraw- 
Hill Book Co., 1948, Chapters 12 and 13. 

9. "Industrial psychology pays in this plant," Modern Industry, (July 15) 1948, 
16, 67 ff. 

10. E. Jaques, "Studies in the social development of an industrial community (The 
Glacier Project), 1," Human Relat., 1950, 3, 223-51. (Sec also E. Jaques, The 
Changing Culture of a Factory. London: Tavistock Publications, 1951 .) 


11. D. Katz, "Morale and motivation in industry" in Current Trends in Industrial 
Psychology. University of Pittsburgh Press, 1949, 145-71. 

12. J. Levine and J. Butler, "Lecture vs. group decision in changing behavior/' J. Appl. 
Psychol, 1952, 36, 29-33. 

13. K. Lewin, "The research center for group dynamics/' Sociometry, 1945, 8, 126- 

14. K. Lewin, "Group decision and social change" in T. N. Newcomb and E. L. 
Hartley (Eds.), Readings in Social Psychology. Henry Holt & Co., 1949, 330- 

15. R. P. Lynton, Incentives and Management in British Industry. Routledge and 
Kegan Paul, Ltd., 1949. 

16. N. Maccoby, "Research findings on productivity, supervision and morale" in 
Research on Human Relations in Administration. University of Michigan: Insti- 
tute for Social Research (Mimeographed Report), March 11, 1949. 

17 C. A. Mace, "Advances in the theory and practice of incentives/' Occup. Psychol., 
1950, 24, 239-44. 

18. N. R. F. Maier, Principles of Human Relations. John Wiley and Sons, Inc., 1952. 

19. F. C. Mann and R. Likert, "The need for research on communicating research 
results" in Papers Presented by Staff Members of the Institute for Social Research 
at the Meetings of the American Psychological Association at Denver, Colorado, 
Sept., 1949. University of Michigan, 1949. 

20. A. J. Marrow, "Group dynamics in industry implications for guidance and per- 
sonnel workers/' Occupations, 1948, 26, 472-76. 

21. A. J. Marrow, "Human factors in production," Personnel, 1949, 25, 341-49. 
21a. M. G. Preston, "Methodological considerations" in H. Helson (Ed.), The- 
oretical Psychology. D. Van Nostrand Co., 1951, Chapter 1. 

">0 J3x^,o*,'^Mo *x% I xik^-'e. 7OC? 


The Work Group as an 
Informal Social Organization 

IN A discussion of difficulties arising in collective bargaining, 
Selekman describes an incident which occurred in a men's clothing manu- 
facturing plant (15). Five men joiners, who had worked together for 
years, carried through the complete process of sewing together the five 
sections of the coat. New methods were introduced by the management 
which required each man to sew only one section. Although they were 
all pieceworkers, with varying earning capacities, the five men arranged 
to pool their total earnings at the end of each week and to divide them 
into five equal parts. "They had always been friends, they explained, 
and so wanted to avoid any bad feelings different earnings might create 
among them." In this way, they resisted a program formulated by man- 
agement which would serve to transform the close group into five com- 
peting individuals. Another group, including four women canvas bastcrs, 
developed a comparable device for "defending their old work-group 
integrity against the erosion of the new methods." They, like the joiners, 
differed from one another in speed, accuracy and earning capacity under 
a piecework wage plan. When there was a shortage of work, none would 
work unless each had a garment to sew on. "Time and again," says 
Selekman, "they had held up production because 'friends like us' could 
not let management give too much work to the speedier or better basters. 
So strong was their cohesiveness that when an arbitrator expressed the 
desire for interviews to get to the bottom of complaints, they refused to 
meet him unless they were interviewed as a group. None would talk 
except in the presence of all" (15, p. 29). 

Another illustration of the drive toward social cohesiveness is found 
in a study of attitudes of building operators conducted in Great Britain 



in 1946, by the Building Research Unit of the Medical Research Coun- 
cil. This investigation showed that nearly 60 per cent of the men were, 
on the whole, opposed to wage-incentive plans, i.e. to "payment by re- 
sults." Behind this, to a large extent, was the fear that this incentive 
would inhibit other strong and pleasurable motives for working, in- 
cluding the solidarity and good fellowship of the working groups to which 
the operators attached great importance (4). 


The incidents described above illustrate a fact which appears again 
and again in the studies described in earlier chapters, namely that "the 
people of an organization are not just a bunch of individuals huddled 
together in a mass. They form groups that are social unities" ( 5, p. 17) . A 
most important outcome of experiments during the period starting with 
about 1930 has been to focus attention upon the influence of the worfe 
group upon productivity, motivation, and morale. Of particular signifi- 
cance is the recognition of the peculiar role of the informal social or- 
ganization the spontaneously-organized smaller work group existing 
within the broader formal structure of the industrial plant. 

Historically, the emphasis upon social organization at the work group 
level, although anticipated in the social philosophy of Durkheim (6) and 
systematically developed in the work of Lewin and his associates (Chap- 
ter 6), is most directly derived from the early experiments of Elton 
Mayo (1, 2, 9, 10). The latter was largely responsible for undermining 
the philosophical and popular assumption that "mankind is an unor- 
ganized rabble" and for postulating in positive terms the viewpoint that 
in work, as in other activities, "man's desire to be continuously associated 
. . . with his fellows is a strong, if not the strongest human characteris- 
tic" (11). As elaborated in the publications of Whitehead (16, 17, 18), 
of Roethlisberger, and of the other members of what is commonly 
known as the Harvard Group (14), the perspectives of Mayo have led 
to an emphasis upon the significance of interactions among members 
of a working group and of the reactions of individuals conscious of 
their membership in one group to those representing another group 
, P . 59). 

Studies in the Hawthorne Plant of the 
Western Electric Company 

The nature of such (social) sentiments and feelings and of the routine 
patterns of interaction in the work group has been most extensively 
investigated in experiments conducted in the Hawthorne plant of the 
Western Electric Company (14, 18) during the years 1927-32. The series 


of investigations in the Hawthorne plant started with a study of the 
relations of levels of illumination to levels of output. This showed that 
production rose as the amount of light was increased for an experimental 
group. However, the output of a control group also went up, even though 
no increase in the amount of illumination was provided. Furthermore, 
when the level of illumination provided to the experimental group was 
decreased to a level below that furnished for the control group, output 
continued to rise, as did also that of the control group. 

Such findings suggested that modifications in illumination, with re- 
sulting changes in levels of "fatigue" experienced by the operators, were 
not responsible for the observed variations in production. As a result, 
research was extended to an examination of the effects of other condi- 
tions of work in an effort to identify the variables operating in this 


As a first step, five girls employed in assembling telephone relays were 
segregated from their regular department and formed into a Relay Assem- 
bly Test Group * for special study. The group was established by inviting 
two experienced operators who were known to be friendly to each other 
to take part in the study. They were then asked to choose the remaining 
members of the experimental group. Following their introduction to 
the test room, conditions of work were systematically varied, in ac- 
cordance with the schedule shown in Table 15, with the view, originally, 
of examining the effects of variables such as the length of the work 
periods, rest pauses, etc., upon production and adjustment 2 (12). 

Records of production were maintained by means of an electric device 
recording the completion of each relay as it passed out of a chute into 
which it was dropped by the operator. As a basis of comparison, a record 
of output for two weeks in her regular department was kept for each 
test operator (without her knowledge) before she was moved to the test 
room. After being moved to the test room no change in conditions of 
work was made for a period of five weeks, although a record of output was 
kept during this period. 

The first innovation in conditions of work (Period 3) was a change 

1 The group actually consisted of six girls, including five engaged in the assembly 
operation and the sixth acting as a layout operator who assigned work and procured 

2 In the course of the study, data were also collected on the amount of sleep, recrea- 
tions, home conditions, and observations were also made of the workers in the plant. 
Pulse rate, blood pressure, blood condition readings, vascular skin reactions, and 
other physical examination records were obtained from time to time. Such data were 
considered in the evaluation and interpretation of changes in production records and 
in arriving at the conclusions discussed in this section. 


in method of payment, the operators being placed on a group piece-rate 
base salary independent of the large group with which they had previ- 
ously been identified. Following this, a series of changes in working con- 
ditions, consisting chiefly of variations in the length and position of rest 
pauses, as shown in Table 15, was introduced. 

^ 2800 






I' 1 "" 1 litlllllllllllllimilll 


/ 2 J 1 S 6 




Figure 15. The Performance Curves for Two Operators in the Relay As- 
sembly Test Group.* Periods of the test are separated by the vertical lines. 
The test conditions in each period are shown in Table 1 5. 

In Figure 15 are presented the production records of two operators 
throughout the period of the experiment. It is evident from these data, 
which are representative of those obtained for the five subjects, that the 
productivity of the test group tended to increase regardless of the nature 
of the changes in working conditions introduced, the average increase 
attained by the end of the period being about 30 per cent. Moreover, 
this trend continued even during the 12 weeks of Period 12 when the 
conditions of work characterizing the regular relay assembly department 
were reintroduced, namely, a full 48-hour workweek with no rest periods 
and no midmorning refreshments.* However, as can be seen from Figure 
15, production rose to new and record heights in the subsequent 31-week 
period (Period 13), when rest pauses were again allowed. 

In setting up the Relay Assembly Test Group, it had been hypothesized 
that changes in specific work conditions, such as the wage-payment 
plan, hours of work, the introduction of rest pauses, etc., would be ac- 
companied by related changes in production such as had been reported 

1 Week of 5% days; 8% hours for 5 days and 4^ hours on Saturday. 

* By permission, from G. A. Pennock, "Industrial research at Hawthorne," Pers. /., 



TABLE 15. Summary of Test Periods, Relay Assembly Group (after Pennock) 

number Period name 

Dates included tion 

Beginning of Rest Pauses 

A. M. P. M. 




In regular department 

4-25 to 5-10 




Introduction to test room 

5-10 to 6-11 




Special gang rate 

6-13 to 8- 6 




Two 5-minute rests 

8- 8 to 9-10 





Two 10-minute rests 

9-12 to 10- 8 





Six 5-minute rests 

10-10 to 11- 5 


8:45,10:00, 2:00,3:15, 




1 5-minute A. M. lunch 

11- 7 to 1-21-28 




10-minute P. M. rest 



Same as No. 7, but 4: 30 


1-23 to 3-10 





Same as No. 7, but 4:00 


3-12 to 4- 7 





Same as No. 7 (check) 

4- 9 to 6-30 





Same as No. 7, but Sat- 

7- 2 to 9-1 




urday A. M. Off 


Same as No. 3 (no 

-*- 3 to 11-24 



lunch or rests) 


1 3 Same as No. 7, but op- 

erators furnish own 
lunch. Company 
furnishes beverages 

14 Same as No. 11 

1 5 Same as No. 7, except 

operators furnish own 
lunch. Company fur- 
nishes beverages 

11-26-28 to 6-29 31 


to 8-31 
to present 





* Approximately. 

in earlier studies on motivation, fatigue, and monotony at work. The 
consistent rise in output throughout the experiment even when rest 
pauses were eliminated and hours of work lengthened led the investi- 
gators "to change their ideas radically." As they reviewed the course of 
the experiment, they came to the "realization" that in the establishment 
of the test room almost all the practices common to the shop had been 
altered. In order to enlist co-operation in the experiment, the operators 
had been consulted about the changes to be made. Careful considera- 
tion had been given to their reactions to the changed conditions as as- 
certained in conferences which, at times, took place in the office of the 


superintendent. Contrary to the practice in the department from which 
they were drawn, the girls were allowed to talk at work. The investigators 
had displayed concern for their health and well-being and interest in 
their opinions, their fears, their hopes and aspirations. Consideration 
of these facts, in relation to the consistently rising production record 
(particularly as seen in Period 12) led to what Roethlisberger has called 
the great eclaircissementan enlightenment quite different from what 
had been expected (13, p. 13 f.). This change in orientation has been 
colorfully described by Stuart Chase in a review of the Hawthorne study 

"The staff swooned at their desks. They had thought they were re- 
turning the girls to original conditions but found that those original 
conditions were gone forever. Because of some mysterious X which had 
thrust itself into the experiment, the experiment had changed under 
them, and the group they now had was not the group they had started 

"This X wasn't in the production end of the factory. It was in the 
human end. It was an attitude, the way the girls felt about their work 
and their group. By asking their help and co-operation, the investigators 
had made the girls feel important. Their whole attitude had changed 
from that of separate cogs in a machine to that of a congenial group 
trying to help the company solve a problem. They had found stability, 
a place where they belonged, and work whose purpose they could clearly 
see. And so they worked faster and better than they ever had in their 
lives" (3, p. 17). 


This "claircissement" led to a new series of investigations, designed 
to check on certain details of the Relay Assembly Room study, and also 
to test the assumption that the observed production increase was a re- 
sult of a change in the social situation that employees did their best, 
not primarily because of wage incentives, reduced fatigue, or similar 
factors, but because the (new) social situation was conducive to maxi- 
mum co-operation by the group in carrying on the work of the unit. 

As employees in their regular department, the operators in the Relay 
Assembly Test Group had participated in the earnings of a group of ap- 
proximately 100 employees. Since the group was large, the output of 
any worker was not directly reflected in her weekly wages. This group- 
payment plan was supplemented by an additional financial incentive 
in the form of a "bogey" system under which a rating of each worker's 
performance was used as a basis for determining the individual's hourly 
rate of pay. 

In Period 3 of the experiment the wage payment plan was altered in 
two respects: viz., (1) the workers in the test room were formed into 
a separate group for purposes of payment, and (2) the effect of the 


"bogey" was nullified by telling the operators to work at a comfortable 
and natural pace. According to Roethlisberger and Dickson (14), it 
seemed possible that the "bogey" continued to function in that each 
operator still kept in mind a standard of what constituted a "good day's 
work." Moreover, the change from the large group to the small one re- 
sulted in an enhancement of the monetary incentive, since each operator's 
earnings became more directly related to her own production. This 
situation, together with the fact that the output of the Relay Assembly 
Test Group continued to rise regardless of the nature of other changes 
in working conditions, suggested the need for examining, among others, 
the hypothesis that the continued increase in output throughout the 
two-year period reflected the continued effect of the wage incentive, in 
other words, the hypothesis that the operators were primarily motivated 
by economic factors. 

As a result, two additional studies were initiated. The aim of the first 
of these experiments was to reproduce the original test room situation 
only in respect to the one factor of method of payment. Underlying 
this was the thought that any observed changes in output could be 
reasonably related to this factor, since the modification of the wage plan 
would represent the only change from the usual conditions of work. In 
the second experiment, the test room situation was to be duplicated in 
all respects except for the change in the wage plan. In this case, it was 
postulated that an increase in output similar to that observed in the Re- 
lay Assembly Test Room would suggest that the enhancement of the 
wage incentive was not the dominant factor in the situation. 


The first of the supplementary experiments involved a group of opera- 
tors known as the Second Relay Assembly Group, consisting of five ex- 
perienced relay assemblers selected by the foreman of the regular de- 
partment and formed into a special group to be paid separately from 
the rest of the department. Before the study began, the five operators 
had been scattered throughout the assembly group. For purposes of the 
study and to facilitate recording of output, they were moved to adjacent 
positions at a bench in the regular department. No other changes were 
made, supervision and general working environment remaining the same. 

The study provided an opportunity for the comparison of output dur- 
ing three periods: 

1. a control period of five weeks during which the wage plan of the 
regular department remained in effect; 4 

4 In planning the experiment it had been the intention of the investigators to take 
as a base, against which subsequent output changes could be measured, the output 


2. an experimented period of nine weeks during which the method 
of payment was altered to coincide with that introduced in the original 
test room in Period 3; 

3. a control period of approximately seven weeks marked by a return 
to the original wage payment system. 

Table 16 shows the average hourly output by weeks for each of the 
operators in the Second Relay Assembly Group during the three periods. 
The increase in output during the period when the wage incentive was in 
effect, followed by a production decrease with the elimination of the 
wage incentive, represents evidence ordinarily interpreted as indicative 
of the direct and favorable influence of financial incentives upon output, 
and of the operation of economic motives in the industrial situation. 
However, as appears on page 205, the investigators contend that conclu- 
sions concerning the efficaciousness of the wage incentive system would 

TABLE 16. Average Hourly Output in Each Period and Percentage Change 

in Average Hourly Output from Period to Period * 

(after Roethlisberger and Dickson) 

Second Relay Assembly Group 

After return 

Experimental to old method 

Base period period of payment 

Operator Average Average Average 

Hourly Hourly Hourly 

Output Percent Output Percent Output Percent 


































* The method of recording output in the Second Relay Assembly Group 
was different from that employed in the Relay Assembly Test Room. Instead 
of expressing the output in terms of the number of relays assembled, it was 
expressed in terms of the number of relay parts assembled. This method, 
although not so accurate as the method used in the test room, eliminated the 
difficulties encountered in attempting to convert the output figures to a com- 
mon basis. 

records for a period of one or two weeks immediately before the change in wage in- 
centive. For various reasons, the base period record was selected from a period having 
similar overtime conditions; it covered five weeks from August 27 to September 29, 
1928. In the case of Operator Ri and Operator R, output records for only two weeks 
and one week respectively during this period were obtainable, as they were doing 
miscellaneous work at the time. 


be entirely misleading without a consideration of the social situation 
characterizing the Second Relay Assembly Group. 


In the second supplementary experiment, involving a Mica Splitting 
Test Group, the objective was to create a test-room situation wherein 
changes in working conditions similar to those imposed in the original 
Relay Assembly Test Room could be introduced without changing the 
wage-incentive method. However, two secondary objectives were con- 
sidered in planning the Mica Splitting Test Room. Since the short dura- 
tion of each of the experimental periods had rendered an interpretation 
of the results in the Relay Assembly Test Room difficult, the periods in 
the Mica Splitting Test Room were to be lengthened to provide for "com- 
plete adjustment to the experimental change before a new condition was 
imposed" (14, p. 134). Furthermore, it was decided to study the effects 
of overtime. 

Earnings on the mica-splitting job were relatively high and, in general, 
the job was considered one of the most desirable of those available for 
women in the Hawthorne plant. Five experienced mica splitters were 
selected from the regular department in accordance with the procedure 
used in selecting subjects for'the original Relay Assembly Test Room. 
In order to furnish a base record with which subsequent changes could 
be compared, output data were collected for a period of eight weeks for 
each operator (Period 1 ), while the girls were still working in the regular 
department. The operators were then isolated from their regular work 
group and placed in a separate room where conditions, in general, dupli- 
cated those characterizing the original Relay Assembly Test Room experi- 

During the first five weeks in the test room, constituting Period 2, no 
changes were introduced, and overtime hours which had been in force 
for some time in the regular department were continued. This meant that 
the operators worked 55 hours per week, including 7% hours on Saturday. 

In Period 3, which lasted 29 weeks, two ten-minute rest pauses were 
instituted, one at 9:30 A.M. and the other at 2:30 P.M. Overtime work 
was continued and on seven occasions Sunday work was added. 

During Period 4, which extended for 48 weeks, overtime hours were 
discontinued and the operators were returned to the standard 48 hour 
working week of 5% days, 8% hours on five days, with 4% hours on Satur- 
day, but the rest pauses were continued. 

During Period 5, the final change in working conditions was made, 
not as an experiment but rather because of the general business depres- 
sion. The length of the working day was reduced to eight hours, Satur- 
day work was eliminated entirely, although the rest pauses were continued. 


These conditions were in effect for 17 weeks, i.e. until the study was 
terminated on September 13, 1930, when the mica-splitting job was dis- 

In general, the changes in each of the last three periods represent im- 
provements in working conditions as compared with preceding periods. 
According to Roethlisberger and Dickson, in view of these successive 
improvements, it might be expected that average output would rise 
steadily. In fact, as shown in Figure 16, the introduction of rest pauses 


Per " 


Au$. Oct. 
27 n 1 

3 ExperiwenUl Periods 4 



T r 



i^fc* I isiq l 1930 

Figure 16. Average Hourly Output per Week, Mica-Splitting Test Room * 

* By permission, from F. J. Roethlisberger and W. J. Dickson, Management and 
the Worker, Harvard University Press, 1939. 


in Period 3 was accompanied by an immediate and sustained rise in the 
weekly average hourly output for each operator, with Operator M-5 as 
a possible exception. This continued during the first few months of Pe- 
riod 4 when, in addition to continuing the rest pauses, hours of work 
were reduced to the standard 8% hour day by the elimination of over- 
time. This increase in output, representing an average rise of 15 per cent 
in the first 14 months of the experiment, would ordinarily be accepted 
as evidence that the introduction of rest pauses and the shortening of 
the work day can in themselves result in increased output, even in the 
absence of changes in the way of enhancing the wage incentive. 

While this possibility is suggested by those concerned with the inter- 
pretation of the Hawthorne data, emphasis is also placed on the decline 
in average hourly output which started after the first few months of 
Period 4, continued for the remainder of this period and also through 
Period 5, in spite of the reduction of the work week to 40 hours. Accord- 
ing to the investigators, this decline, extending through approximately 
the second year of the experiment, can be more properly related to the 
workers' anxieties arising over the uncertain future of the mica-splitting 
job which started when all mica operators in the regular department were 
removed to a new location in the plant and transferred to a new depart- 

Such anxieties became more and more severe as other changes were 
made and came to a climax with the realization that "the dreaded de- 
pression had begun to affect them" when, on May 19, 1930, the hours of 
work were reduced and Saturday work eliminated. It is suggested that 
logically, in this situation, it might have been expected that the monetary 
incentive would be at the peak of its efficiency, that operators would 
have wanted to earn as much money as possible before being laid off. 
However, output dropped. Fears and anxieties completely overbalanced 
the experimentally introduced changes; in the face of such doubts the 
wage incentive had lost its power to motivate. 

According to Roethlisberger and Dickson, although this unexpected 
factor had interrupted the experiment proper, it brought out certain 
points which were of value in showing the effect of interfering preoccupa- 
tions on the attitudes of the operators and, in turn, on their output. This 
unanticipated condition, it is said, demonstrated that under certain con- 
ditions such preoccupations could become so acute as to nullify the ef- 
fects of certain other factors influencing output and efficiency. It pro- 
vided a further warning against drawing hasty conclusions based on only 
a partial acquaintance with the human situation under investigation ( 14 
p. 154). * V ' 

The fact that the danger of losing the job may lead to a decrease in 


production, with the view of "stretching work/' even in the face of strong 
wage incentives, particularly in the absence of pressure for production by 
plant management, has been established in the German Industrial In- 
quiry (7, 8) and other studies. For this, and other reasons, it is not dif- 
ficult to accept the viewpoint that attitudes arising in the work situation 
affect the operation of financial incentives. However, the rapporteurs 
of the Hawthorne studies appear to neglect other conditions which need 
to be examined objectively in evaluating the significance of the financial 
incentive itself as a stimulus to response in the way of attitudes and be- 
havior. So, for example, note is made of the particularly marked decline 
in output during the final six months of the Mica Splitting Room study 
which is evident from an examination of Figure 16. Attention is also 
drawn to the fact that during this period operators were working on mica 
of different specifications, necessitating a conversion of output to one 
standard type of mica. It is pointed out that both "the investigators and 
operators were of the opinion that the rates on the new piece parts were 
not high enough in comparison with the old." Nevertheless, scant con- 
sideration is given to the possibility that the discrepancy in the piece rates 
resulted in a weakening of the financial incentive; that a reduced appeal 
to economic motives could readily account in large part for the very severe 
drop in output observed during this final phase of the Mica Splitting 
Room experiment 


According to Roethlisberger and Dickson, although output had risen 
on the average 12 per cent in the Second Relay Assembly Group study, 
factors other than the change in wage incentive contributed to this in- 
crease. Since the operators were not segregated in a separate room but 
had remained in their regular department, it had been assumed that they 
were a "group" in name only. Actually, a new variable which had not 
been considered important in designing the experiment had entered into 
the situation. The operators had not been left in their original positions 
at different benches, but had been moved to a common bench in the 
regular department in order to facilitate record keeping. This move, it is 
claimed, had social significance in bringing into focus a competitive 
attitude with respect to the original Relay Assembly Test Group already 
existing in the department. 

"There was some evidence," it is said, "to indicate that the operators 
in the Second Relay Assembly Group had seized upon this test as an 
opportunity to prove to everyone that they could do as well as the Relay 
Assembly Test Room operators. They were out to equal the latter's record. 


In view of this, even the most liberal estimate would put the increase in 
output due to the change in payment alone at somewhat less than 12 
per cent (14, p. 158). . . . "This experiment, designed to test the effect 
of a single variable, succeeded only in exposing a most complex social 
situation. Conclusions about the efficacy of a wage incentive drawn from 
it, unrelated to the basic social situation, would have been entirely mis- 
leading" (14, p. 577). 

In the Mica Splitting Test Room, too, according to Roethlisberger 
and Dickson, there had been a similar failure to consider important varia- 
bles representative of changes in "social atmosphere/' The experimental 
changes imposed in the original Relay Assembly Test Room not only 
differed from those introduced in the Mica Splitting Test Room in char- 
acter, duration, and frequency, but they functioned for the two groups 
in a different way. Those introduced in the original Relay Assembly Test 
Room were in the nature of special privileges while those introduced in 
the Mica Splitting Test Room, except for the rest pauses, were similar 
to changes in the regular Mica Splitting Department. Mica splitters, who 
were on a straight piecework rate, did not have a vital interest in one an- 
other's output around which to organize as a social group. It is further 
assumed that operators would tend to work closer to their maxi- 
mum capacity under a straight piecework than under a group piecework 

On the basis of such considerations, the conclusion is reached that, 
apart from the wage factor, conditions in the Mica Splitting Test Room 
were not so conducive to an increase in output as they were in the Relay 
Assembly Test Room. Nevertheless, in spite of this, average output in- 
creased 15 per cent during the first 14 months of the Mica Splitting Test 
Room study, prior to the appearance of anxieties with respect to job 
security. If, according to Roethlisberger and Dickson, this is allowed as 
the minimum due to changes in working conditions and supervision, i.e. 
all factors other than "wage incentive, then this amount subtracted from 
the percentage increase in output in the original Relay Assembly Test 
Room should leave the maximum amount to be attributed to the change 
in wage incentive. Since output rose an average of roughly 30 per cent in 
the original Relay Assembly Test Room, there remains 1 5 per cent which 
"might be attributed to the change in wage incentive/' 'This deduction/' 
it is pointed out, "depended on so many assumptions which might or 
might not hold true" that it failed in any sense to be conclusive. With 
respect to over-all outcomes, Roethlisberger and Dickson take the posi- 
tion that at least two conclusions seemed to be warranted from the test 
room experiments 


"1. There was absolutely no evidence in favor of the hypothesis that 
the continuous increase in output in the Relay Assembly Test Room 
during the first two years could be attributed to the wage incentive fac- 
tor alone; 

"2. the efficacy of a wage incentive was so dependent on its relation 
to other factors that it was impossible to consider it as a thing in itself 
having an independent effect on the individual. Only in connection with 
the interpersonal relations at work and the personal situations outside 
of work, to mention two important variables, could its effect on output 
be determined" (14, p. 160). 

As has been suggested in the discussion of the two supplementary ex- 
periments, questions can be raised concerning the interpretation of find- 
ings. In particular, the experiments do not demonstrate that rest pauses 
and wages are without value as incentives to production. Furthermore, 
they do not justify the firm conclusion that these (or other conditions) 
fail to exercise independent effects upon the individual. Nevertheless, 
the experiments served an important purpose in calling attention to the 
fact that interpersonal relations and the character of the social situation 
can alter the effects of such specific incentives. The studies also point 
to the need for the detailed examination of the structure of the social 
situation at the level of the work group, and of the interrelationships 
among those specific factors environmental and personal which may 
determine the attitudes and behavior of the individual as a member of a 
work group. While questions of the type discussed in this section can be 
raised, it nevertheless seems clear that these early studies in the Hawthorne 
plant were highly significant (1) in calling attention to the existence 
and importance of "social" motives in industry, and (2) in showing the 
need for intensive investigation of the pattern of group activity of the 
functions and structure of the "team" as represented in studies of the 
type discussed in Chapter 11. 


1. C. M. Arensberg, "Behavior and organization: industrial studies" in J. H. Rohrcr 
and M. Sherif (Eds.), Social Psychology at the Crossroads. Harper and Brothers, 
1951, 324-54. 

2. R. Bendix and L. H. Fisher, The Perspectives of Elton Mayo (Reprint No. 17). 
University of California, Institute of Industrial Relations, 1950. 

3. S. Chase, "What makes the worker want to work/' Reader's Digest, (February) 
1941, 38, 15-20. 

4. N. M. Davis, "Attitudes to work among building operatives," Occup. Psycho/., 
1948, 22, 56-62. 

5. R. Dublin, Human Relations in Administration. Prentice-Hall, Inc., 1951. 


6. E. Durkheim, De la Division du Travail Social, 1893. (Sec Division of Labor in 
Society, Translated by G. Simpson. Glencoe, 111.: Free Press, 1947.) 

7. O. Lipmann, "Arbeitswissenschaft," An Bet. und Arb., 1929, 3, 76-82. 

8. O. Lipmann, "Ergebnisse der Arbeitsforschung/' An. Bet und Arb., 1929, 3, 178- 

9. E. Mayo, "Revery and industrial fatigue/' Pers. /., 1924, 3, 273-81. 

10. E. Mayo, The Human Problems of an Industrial Civilization. Harvard Business 
School, Division of Research, 1946. (First printed by The Macmillan Co., 1933.) 

11. E. Mayo, The Social Problems of an Industrial Civilization. Harvard University 
Press, 1945, Chapter II. 

12. G. A. Pennock, "Industrial research at Hawthorne/' Pers. /., 1930, 8, 296-313. 

13. F. J. Roethlisberger, Management and Morale. Harvard University Press, 1939. 

14. F. J. Roethlisberger and W. J. Dickson, Management and the Worker. Harvard 
University Press, 1939. 

15. B. M. Selekman, "Living with collective bargaining/' Harvard Bus. Rev., 1941, 
22, 21-33. 

16. T. N. Whitehead, Leadership in a Free Society. Harvard University Press, 1936. 

17. T. N. Whitehead, "Social motives in economic activity/' Occup. Psychol., 1938, 
12, 271-90. 

18. T. N. Whitehead, The Industrial Worker. Harvard University Press, 1938. 




"AN INSTITUTION," according to Drucker, "is like a tune: 
it is not constituted of individual sounds but by the relations between 
them" (7, p. 26). This comparison applies not alone to the large and 
formally structured organization, but also to the small and more spon- 
taneously organized work group. At least, such is the conclusion which 
has been drawn from the study of interpersonal relations in the small work 
group and of the patterning of "teamwork" in industry. 

Bank Wiring Observation Group Hawthorne 
Branch of the Western Electric Company 

The series of studies conducted in the Hawthorne plant of the Western 
Electric Company, described in Chapter 10, included a detailed study 
of a shop situation undertaken with the purpose, in part, of examining 
social interactions in work groups in relation to the operation of wage 
incentive plans. In studies conducted early in 1931, attention had been 
called to the fact that social groups in shop departments were capable 
of exercising very strong control over the work behavior of their individual 
members. Observations made in the course of an extended interviewing 
and counseling program, described in Chapter 13, suggested that the 
wage-incentive systems under which some of the groups worked had been 
rendered ineffectual by group pressure for restriction of output. Informal 
practices by means of which operators were placed under pressure were 
brought to light, as well as evidence of informal leadership on the part 
of some individuals who assumed the responsibility of seeing "that the 
members of a group clung together and protected themselves from rep- 
resentatives of other groups within the company who could interfere 
with their affairs" (22, p. 380). 

Such observations led to the Bank Wiring Observation Room study. 



This involved a group of 14 male operators who, in order to facilitate 
record keeping and observation by investigators, were placed in a sep- 
arate room. 1 This group included nine wiremen arranged in three sub- 
groups of three men each. Six of the wiremen, i.e. two subgroups, worked 
on "connector" equipment and the remaining three on "selector" equip- 
ment. One solderman was attached to each of these subgroups, con- 
stituting a soldering unit. Two inspectors passed on the work of these 
12 men. In addition to the 14 operators, there was a supervisor in charge 
of this "small shop." The study lasted for a period of six and one-half 
months (November, 1931 to May, 1932), being terminated at the end 
of that time because of lack of work. No intentional changes were made 
in the situation once the group was so placed that observation was pos- 

Contrary to the procedure followed in the earlier studies, in order to 
assess the effect of placing the group in a separate room, base-period 
studies were made before either the workers or their immediate super- 
visors knew anything about the study. It was decided nothing was to be 
said or done in selecting the group, in explaining the study to them, or 
removing the men from the department which might alter their status 
in any way (22, p. 388) . In addition, in planning the study it was agreed 
that no records were to be taken which might tend to make the workers 
apprehensive or too consciously aware that they were being studied. 
Nevertheless, provisions were made for obtaining records on quantity of 
output, which were accumulated by the observer, for each of the nine 
wiremen, at noon and at night after the men had quit work. 

Records on quality of output were also obtained daily, and the ob- 
server maintained a daily record of observations, including significant hap- 
penings in the observation room, conversations of the employees, and his 
own impressions. In addition, with the exception of the inspectors, each 
man in the group was interviewed before the study began, and all the men 
were interviewed two or more times during the study itself. In connection 
with certain analyses, use was made of physical examination records and 
of records of results on a battery of mental and dexterity tests, which were 
applied toward the close of the study. 

Wages paid to the operators were based on a system of group piecework 
according to which the entire department was considered a unit for pur- 
poses of payment. Under this system the earnings of any one individual 
were affected by the output of every other member in the group, and the 
wages of an individual worker who turned out an unusually large amount 

1 The group of 14 men originally selected for the study did not remain intact very 
long, and substitutions were made from time to time in the case both of operators 
and of supervisory personnel. 


of work, in comparison with the output of other workers, would be but 
slightly higher than if he had not increased his output at all. Conversely, 
the output of the other operators would serve to sustain at about their 
ordinary level the wages of a worker who was unusually low in output for 
a time. To help minimize inequities of this type, adjustments in hourly 
rates of the operators, which determined the "daywork value" of each 
operator's work, were based largely upon individual outputs, records being 
kept of each person's production, in terms of weekly average hourly out- 
put, for guidance in adjusting rates. Under this system of payment, differ- 
ences in the earnings of different operators working the same number 
of hours depended entirely upon differences in individual hourly rates. 


A major finding of the study, involving detailed analysis of the records 
on quantity and quality of output and of approximately 600 typed pages 
of observation and interview data, was that each individual in the group 
-was restricting his output. The restriction of output was manifested first 
in the fact that the group's standard of a day's work was lower than that 
of the "bogey" set up in the wage-incentive plan, the group's standard 
serving to fix an upper limit of each operator's output. Restriction of out- 
put was also evidenced, according to the investigators, in the fact that 
each individual's output, as illustrated in sample records in Figure 17, re- 
mained fairly constant from week to week, producing departmental out- 
put curves which were devoid of individuality and approximated horizon- 
tal lines in shape. A comparison of average hourly output during morning 
and afternoon hours showed further that ( 1 ) the average morning rate 
of each operator exceeded the average afternoon rate and that (2) the 
difference between morning and afternoon rates was greater among faster 
than among slower operators. This, according to the investigators, repre- 
sented further evidence of a concern with the maintenance of uniform 
output in the interests of output restriction. 2 

Although the group chief was supposed to make individual output 
counts at the end of each day, he actually depended upon an output report 
furnished by the operator, since making individual counts would have 
been "a laborious task and an almost impossible one considering his many 
duties." Output counts by the investigators showed, as appears in Figure 
17, that no wireman reported exactly what he produced each day, but 

* The possibility that fatigue or monotony might account for such variations be- 
tween spells of work or variations between operators in work curves receives little 
consideration. Those concerned with the analyses of earlier studies are generally com- 
mitted to the viewpoint that fatigue and monotony, as ordinarily conceived, play 
but a negligible (if any) role in accounting for output decrements on semiskilled 
repetitive jobs of the type involved in this study. 










MOVA4 OEC.ll JAW.* ft*.* MAtt.6 AW.t 

\931 I 

Figure 17. Actual vs. Reported Average Hourly Output per Week, Bank 
Wiring Observation Room * 

reported more on some days and less on others. Such inaccuracies ap- 
peared to be associated with the preoccupation of operators with quantity 
of output, or, more specifically, with the limitation of output. 

The department permitted employees to claim daywork for unusual 
stoppages beyond their control. Observation showed that most of the 
wiremen frequently claimed to have been prevented from working by 
stoppages beyond their control when there was little justification for their 
claims since ( 1 ) the stoppage was shorter than claimed; (2) it was brought 
about by the operators themselves; (3) there was, in fact, no delay, or 
(4) there was a real stoppage which could have been compensated for by 
working a little harder or more consistently. 

Observations and interviews furnished additional evidence of limita- 
tion of output. Most of the operators stated that they could easily turn 
out more work than they did. The observer reported that all men stopped 

* By permission, from F. J. Roethlisbcrger and W. J. Dickson, Management and 
the Worker, Harvard University Press, 1939. 


before quitting time and that those who ranked highest in output stopped 


Various possible interpretations have been postulated and explored 
by Roethlisberger and Dickson in an effort to explain the restriction of 
output and related phenomena in the Bank Wiring Observation Room. 
Hostility to management is considered to represent an untenable explana- 
tion since, according to the investigators, in employee interviews con- 
ducted in the plant in 1929, when over 40,000 complaints were voiced, 
''there was not one single unfavorable comment expressed about the 
company in general" (22, p. 536). Fear of rate cutting is discarded be- 
cause it was the policy of the Western Electric Company not to change 
the piece rates except when there was a change in manufacturing proc- 
esses. In this connection, it is to be noted that such fear may exist in spite 
of the fact that a company refrains from cutting rates as earnings increase. 
Furthermore, John Mills, a former member of the personnel staff of the 
Bell Telephone Company, has suggested that there were factors in the 
bank-wiring situation which were not observed by the investigators. Ac- 
cording to Mills 

"Reward is supposed to be in direct proportion to production. Well, 
I remember the first time I ever got behind that fiction. I was visiting 
the Western Electric Company, which had a reputation of never cutting 
a piece rate. It never did; if some manufacturing process was found to 
pay more than seemed right for the class of labor employed on it if, in 
other words, the rate-setters had misjudged that particular part was re- 
ferred back to the engineers for redesign, and then a new rate was set on 
the new part" (19, p. 93). 

Involved here may be a question of opinion concerning Western Elec- 
tric Company practices in the administration of wage-incentive systems. 
Nevertheless, an "outside" opinion deserves consideration, particularly 
in view of the possibility of errors in observation suggested in the astound- 
ing report by interviewers that no single worker in the Hawthorne plant 
had expressed an unfavorable opinion about the company in general. Of 
further interest is the fact that the study was conducted in the midst of 
a severe depression. The observers reported that in their daily conversa- 
tions operators "speculated endlessly upon when the depression would 
end, whether they would be laid off, and what would happen to them 
if they were. All but one of them were in a very poor financial condition 
and if they were unemployed could not escape public support for long" 
(22, p. 531 ) . As stated elsewhere, other studies have shown that attitudes 
unfavorable to output and a definite effort to "stretch work" may result 


under such conditions, even in the presence of powerful wage incentives. 1 
Nevertheless, it is reported that the interviewers had detected the problem 
of restriction in output even in 1929, the year in which the company 
reached its peak of activity, and the conclusion is reached (contrary to 
that following from the Mica Splitting Room study 4 ) that it is doubtful 
that the restriction of output in the Bank Wiring Observation Room was 
related to the effects of the depression. Furthermore, since all operators 
were found to be physically fit, and no relations found between output 
and performance on mental and dexterity tests, it could not be assumed 
that the failure to meet production standards was related to lack of ca- 

The facts presented in reports on this study suggest the distinct possi- 
bility that the wag^ incentive plan itself and the manner in which it was 
administered were not conducive to optimal motivation. However, this 
possibility in the way of an interpretation receives scant consideration 
by the rapporteurs of the Hawthorne study. Other considerations are also 
laid aside in favor of the conclusion that restriction of output and similar 
problems in the Bank Wiring Observation Room arose from the existence 
of a patterned, informal and intricate social organization developed spon- 
taneously and quite unconsciously in the group through the years with 
the function of resisting change (22, p. 548) or, more specifically "(1 ) to 
protect the group from internaT indiscretions, and (2) to protect it from 
outside interference" (22, p. 523) . The workers, according to Whitehead, 
"were dumbly fearful for their own way of life, and they were jealously 
guarding it from outside interference" (31, p. 78). From this viewpoint, 
to state the failure to meet production standards in terms of "restriction," 
or to state this and other industrial personnel problems as "faulty super- 
vision," or "mismanagement," is to mistake symptoms for causes and to 
neglect the social factors involved. 


The Bank Wiring Observation Room study included an analysis of 
the social structure of the group and the mechanisms used in enabling the 
informal social organization to fulfill its function. This involved an appli- 
cation of the "sociological-anthropological" approach inherent in the per- 
spectives of Mayo and as represented in theories and research discussed in 
Chapter 5. The methods and findings of the study also have significant 
implications in terms of Lewin's field theory, discussed in Chapter 6, and 
in relation to the use of sociometric techniques for graphically portraying 

See page 191. 
* See page 190. 


and analyzing interpersonal relations found in the work of Moreno (20. 


Observation and interview materials were examined for evidence of 
social organizations, to answer the questions: "Do we have here just so 
many 'individuals, 9 or are they related to one another in such a way that 
they form a group or configuration? If they do form a configuration, how 
are they differentiated from or integrated with other groups?" ( 22, p. 493) . 
In seeking answers to these questions each person entering into the study 
was considered separately with the view of discovering the degree and 
kind of his social participation in the Bank Wiring Observation Room. 
The material covering the operator was then examined to provide answers 
to the questions, "To whom do this person's relations extend?' 9 and "Does 
he enter a great deal or relatively little into social relations with the people 
with whom he associates?' 9 

Analysis of findings showed that members of the group were differen- 
tiated or integrated in terms of participation in games, one of the social 
activities observed and recorded in the course of the study. While there 
was no written rule to this effect, helping one another on the job was 
forbidden. However, operators in the Bank Wiring Observation Room 
did help one another even when technically there was no justification for 
doing so. Records kept of this activity showed that everyone 5 participated 
in helping and that, unlike participation in games, helping was not con- 
fined within smaller work groups. Helping one another seemed to in- 
tegrate the whole group rather than parts of it (22, p. 506). 

In addition to studying the direction and frequency of interemployee 
relations, an attempt was made to study the kind of participation mani- 
fested by each operator, whether he assumed a subordinate or superor- 
dinate role; whether he strove for leadership; and whether his moves in 
this direction were accepted or rejected by others. Finally each occurrence 
in which a person entered into association with another was examined to 
see whether the relation manifested expressed an antagonism, a friend- 
ship, or was merely neutral. 

Conclusions have been drawn concerning the nature of the social or- 
ganization in the Bank Wiring Observation Room; its significance as a 
motivating factor in relation to output, and the mechanisms through 
which social pressure was exerted upon individuals in the group for the 
maintenance of group standards and for the effective expression of group 

Examination of the findings showed, for example, that each job in the 
group carried its own social status or significance, that the members of 
the group could be differentialed into five gradations, ranging from high- 

5 Except inspectors, who were not involved in "helping" activities. 



est to lowest in the following order: inspectors, connector-wiremen, 
selector-wiremen, soldermen, and truckers, in terms of social status of 
the job. Further analyses showed that in spite of such differences in job 
status, members of the group did not form occupational subgroups. Nev- 
ertheless, whether the investigators looked at games, job trading, quarrels 
over the windows, or friendships and antagonisms, two subgroups seemed 
to stand out, one located toward the front of the room, the other toward 
the back. "The group in front' 9 and "the group in back" were common 
terms of designation among the workmen themselves. 

The membership of each of these two subgroups or "cliques" is dia- 
grammatically presented in Figure 18. The soldering units into which the 



W S 





Figure 18. The Internal Organization of the Group and Output, Bank Wir- 
ing Observation Room * 

* By permission, from F. J. Roethlisberger and W. J. Dickson, Management and 
the Worker, Harvard University Press, 1939. 


members of the group were divided are shown by the three rectangles, 
and the two large circles demarcate the two cliques, A and B. As indicated 
in the diagram, three individuals, Is, W 5 , and S 2 , were clearly outside 
either clique. The line around W 6 intersects that of Clique B to show his 
partial participation in it, and the instability of WVs position is indicated 
by the broken circle around his number. 

There were differences in the activities of the two groups. Members 
of Clique A did not trade jobs nearly so much as Clique B, and on the 
whole they did not enter into the controversies about the windows. 
Clique A engaged in games of chance, while Clique B engaged more often 
in "binging." 6 Both groups purchased candy from the club store, but pur- 
chases were made separately and neither small clique shared with the 
other. Clique A argued more and indulged in less noise and "horseplay" 
than Clique B. The members of Clique A felt that their conversations 
were on a higher plane than those which went on in Clique B; as W 4 said, 
"We talk about things of some importance" (22, p. 510). 

In general, the members of Clique A regarded themselves as superior 
to Clique B, and the latter occupied the lower position in the over-all so- 
cial structure of the Bank Wiring Observation Room. With respect to 
output, the analysis showed, as appears in Figure 18, that the actual aver- 
age hourly output rates of the members of Clique B were lower than those 
of Clique A and that overestimates of output and claims for daywork 
allowances were greater. In other words, there was a direct correspondence 
between degree of output and informal social standing of the groups. 
Members of Clique A were drawn from the two soldering units working 
on connector equipment, while members of Clique B were drawn from 
the soldering unit working on selector equipment. In the department, 
new men were generally started as selector wiremen and moved "forward" 
to connector work as their efficiency improved. Such increases in efficiency 
were usually rewarded by increases in the hourly rate. Increases in hourly 
rates resulted in increased earnings under the wage-incentive plan. While 
it is suggested that low output might actually have affected the position 
of the group, and that output and group social positions were in a relation 
of mutual dependence, the investigators nevertheless take the position 
that the low output of members of Clique B represented primarily a device 
for "getting back" at those who were displaying their superiority (22, p. 

The two cliques were found to be differentiated in other respects. Each, 

"Binging" refers to ways of expressing personal antagonisms, e.g. in controversies 
about opening or closing windows, through a "controlled" exchange of blows, etc. 
"Binging," according to Roethlisberger and Dickson, was used to help enforce the re- 
striction of output. 


for example, had its own leaders, and procedures for forcing supervisory 
personnel into acceptance of their codes of conduct. Nevertheless, there 
was at the same time considerable solidarity between the two cliques. The 
internal solidarity of the group of the Bank Wiring Observation Group 
as a whole found particular expression in sentiments connected with oc- 
cupation, output, and supervision. These, according to Roethlisberger and 
Dickson, may be summarized as follows 

"1. You should not turn out too much work. If you do, you are a 

"2. You should not turn out too little work. If you do, you are a 'chis- 
eler. 9 

"3. You should not tell a supervisor anything that will react to the 
detriment of an associate. If you do, you are a 'squealer. 9 

"4. You should not attempt to maintain social distance or act offi- 
cious. If you are an inspector, for example, you should not act like one 99 
(22, p. 522). 

Members of Clique A, that is, the people who held the most favored 
position in the Bank Wiring Observation Group, conformed to these 
rules of behavior in all respects. Members of Clique B conformed to rules 
1, 3, and 4, and, it is reported, attached more importance to these rules 
than anyone else. Rule 2, however, was violated by Clique B which, it 
is claimed, indulged in extreme limitation of output as an expression of 
resentment against the social inferiority of the clique. 

The significance for output of such sentiments, associated with the 
informal social organization of employees, has been underlined by White- 
head (31) in a comparison of the original Relay Assembly Test Group 
and the Bank Wiring Observation Group. Each group, he points out, was 
engaged in habitual routines performed in common. Each developed its 
own informal leadership. Both groups grew to have characters and styles 
of performance of their own, and individual standards of conduct were 
adjusted to conform to the group system of ethics. Routines developed 
first as a mere matter of technical convenience or passing pleasure, grew 
into stable customs, and became shot through with sentiments which 
clustered particularly around the economic purpose around the fruits 
of labor as represented by the pay envelope. "Both groups/' according to 
Whitehead, are "examples of the growth of stable human relations the 
way people act together, the sentiments they share together, the system 
of ethics and codes of conduct that regulate these relationships" (31, p. 

There was, however, according to Whitehead, a significant difference 
between the two groups. In the case of the Relay Assembly Test Group, 
behavior and sentiments were so organized as to promote the economic 


purposes of the plant, and this was a matter of real pride to each girl. In 
the case of the Bank Wiring Observation Group, this was not so. Instead 
of collaborating with management, the men were "disguising their activi- 
ties" and were resisting control from above. They derived little pride or 
satisfaction from their work. 

Both groups, adds Whitehead, illustrate the importance of social rela- 
tionships built up in the course of day-to-day activities. The differences 
in sentiments and behavior of the two groups show, he contends, that 
it is not sufficient for management to be active in promoting the eco- 
nomic efficiency and the economic welfare of employees; management 
must also guard and develop their social sentiments (30, 31) . 

The Role of the Work Group as a Source 

of Employee Production, Satisfaction, 

and Morale 

From the viewpoint of the problem of incentives in industry, a major 
outcome of the Hawthorne studies is the conclusion, stated by Roethlis- 
berger and Dickson, that "none of the results gave the slightest substan- 
tiation to the theory that the worker is primarily motivated by economic 
interest. The evidence indicated that the efficacy of a wage incentive is 
so dependent on its relation to other factors that it is impossible to sepa- 
rate it out as a thing in itself having an independent effect" * (22, p. 575) . 
Financial incentives are responded to only in so far as they minister to the 
individual's own social situation. When they are opposed to the trend 
of the stable social organization characterizing a group of workers, they 
lose their power to motivate. Social considerations, according to the 
Hawthorne investigators, also outweigh economic ones in determining 
workers' feelings and attitudes, and thereby in determining the nature of 
individual satisfactions and grievances in the working situation. 

Objections can be raised to these generalizations, particularly to the 
implication that financial incentives cannot have a direct and independent 
influence upon output. As suggested earlier, data from the Hawthorne 
studies which are interpreted as revealing the effect of group sentiments 
can be interpreted as showing the immediate and definite influence of 
a change in the wage plan. Certainly, the findings of the Hawthorne 
studies cannot be accepted, as has apparently been done in some quarters, 
as demonstrating that the worker is not concerned with the size of his 
pay envelope except as an outer symbol of the social value of his job, and 
that he will ordinarily not respond directly with increased effort to an 
enhancement of the financial incentive. 

T Not italicized in the original. 


In his description of quota restriction and "goldbricking" in a machine 
shop, discussed in Chapter 3, Roy considers specifically the assumption 
that restriction in the bank-wiring group was due, as Mayo has suggested, 
to frustrations and an ensuing "lower social code resulting from a lack 
of understanding by workers of the economic logics of management" (18, 
p, 119). Roy takes issue with those who consider "economic man" as an 
entirely fallacious concept. "The operators in my shop," writes Roy, 
"made noises like economic men. Their talk indicated that they were 
canny calculators and that the dollar sign fluttered at the masthead of 
every machine" (23, p. 430) . It could be, he adds, that the operators did 
not exceed their quota because of fear of rate cuts, in other words, pre- 
cisely because they )vere alert to their economic interests. 

In spite of such criticism, the Hawthorne series of studies do furnish 
evidence as do also observations in the plant and other studies discussed 
in this volume that financial rewards do not operate with the same 
strength in all circumstances and do not invariably lead to increased pro- 
duction or better morale. There is, unquestionably, much justification for 
the insistence upon the force of the social situation in determining both 
output and satisfaction at work, and in laying the groundwork for indus- 
trial peace or industrial conflict. There is also much merit in emphasizing 
the complexity of the situation, and of the difficulties and dangers in- 
volved in isolating the influence of specific factors, since it is seldom that 
separate or specific motives function in isolation (15). 

Not alone the Hawthorne studies, but the work of Lewin and his as- 
sociates have made it increasingly evident that the problems of produc- 
tivity, satisfaction, and morale in industry are, at least in part, problems 
of group relationships. It is clear, from other studies described in this 
volume (and also from the Hawthorne studies) that economic motives 
cannot be neglected and that specific factors in the plant situation, such 
as work preference, quality of supervision, knowledge of results, etc., in- 
fluence motivation and morale. It is equally clear, from the Hawthorne 
and other investigations that "it is far too primitive to consider only the 
economic motives of the factory hand . . . Good management has to 
consider the total culture and all aspects of group life" (16, p. 134) . 


Acceptance of this viewpoint focuses attention upon the need for the 
further study of work groups or "teams" with a view toward sketching 
more completely the varieties of such small social organizations; deter- 
mining the factors which give rise to such groups and influence their atti- 


tudes and behavior, and assessing their relation to the effectiveness of 
the larger, formal organization (29). In particular it is necessary to in- 
vestigate in detail the elements in the situation which are conducive or 
detrimental to production, satisfaction, and morale. Of further great 
interest in planning such studies is the question of whether the organiza- 
tion of such groups should be "left to chance" (2) or whether manage- 
ment should take the lead in the structuring of the small work group 
as a necessary step in the accomplishment of the economic and social 
objectives of industry. The need for focusing attention upon this ques- 
tion is particularly apparent at least from management's point of view 
when consideration is given to the frequency with which the primary 
function of the spontaneously-organized small work group appears to be 
that of restricting output. 


Conflicts which develop out of the difference between management's 
perception of the team and the work group's perception of its own or- 
ganization and functions are illustrated in a study reported by French 
and Zander (11). The investigation was conducted in a large business 
office where work was of a routine nature and the morale of employees 
was known to be low. As a first step, a sociometric questionnaire, designed 
to reveal the friendship structure of the office and the forces at work 
toward making friends, was administered to all employees. Analysis of re- 
plies and of other data showed that ( 1 ) there were relatively few friend- 
ship groups in this office; (2) there were great differences in the level of 
production from one friendship group to another; (3) different friendship 
groups had their own group standards, some to work hard and some to 
take it easy; (4) some groups tended to identify themselves with manage- 
ment, and others to "aggress against management." Furthermore, the 
findings showed that the most popular girls were rarely high producers; 
the correlation of popularity in the office with productivity proving to 
be -.67. 

As a result of the transfer of a number of girls to another office, it was 
necessary to break up some of the friendship groups and to change the 
seating of others. From the results of the first part of the study, it was 
predicted that in this reshuffling of the "informal social organization" em- 
ployees with high production records would be rejected by low production 
groups, and that employees with low production would be acceptable as 
friends in any group. Comparisons actually showed a very high negative 
correlation of .85, after transfer, between the level of productivity and 
the change in popularity. Almost every high-producing worker lost popu- 


larity as a result of the shift whereas low-producing workers gained 
friends. 8 According to French and Zander, "it appeared that the small 
friendship groups developed group standards concerning their level of 
production, and they rejected from the group those employees who pro- 
duced too much more than the group standards. The employee was faced 
with the unhappy choice between restricting her production to a low level 
and having many friends, on the one hand, and doing what management 
would call a good day's work and becoming a social isolate on the other" 
(11, P. 76). 

A Study of Teamwork and Productivity 
in a Shoe Factory 

The nature and significance of the conflict between "organization 
goals" and "personal goals" in the work of a team is further illustrated in 
a study by Horsfall and Arensberg (12) of teamwork and productivity in 
a shoe factory. This investigation was conducted in the Bottoming Room 
of the plant in which there were four seven-worker teams, each perform- 
ing the same series of machine operations. The 28 employees constituted 
a single department group, all working in the same room. Each team in- 
cluded two trimmers, a pounder, a rougher, a shanker, and two cementers. 
Production for all workers (except those in training) was on a piecework 
basis. The piece rate unit was a "case" or "rack" of 36 pairs or 72 shoes. 
Earnings under this system were freely compared by workers and were 
common knowledge. 

In addition to the operations peculiar to each specialized job, some 
members of the team handled the racks on which the shoes were pushed 
from one operation to the next. Observations in the Bottoming Room 
showed that this was actively related to the starting of work operations 
in a team. Specifically, this was a method of allocating racks in an order 
which was different from that in which they were received from the Last- 
ing Room of the plant. Furthermore, for each type of shoe there was a 
corresponding object a "coin" of a social economic system which had 
come to serve as a symbol in an unofficial record-keeping and exchange in 
the workers' own routing of racks on a well-defined order of allocation of 
work by the four teams. Observation and interviews showed further that 
"the racks to be worked on had been routed among the teams in a manner 
to equalize pay for fast and slow teams alike, and assure that all the work- 
ers should score a uniform production of racks." Workers explained that 
this system, originally introduced by the roughers, prevented quarrels and 

Exploratory sociometric studies conducted by the Industrial Relations Center, 
University of Minnesota, have also yielded findings (characterized as tentative) which 
indicate that employees do not choose the best producers as their best friends (26). 


provided assurance that no one team would get all the "hard" racks, al- 
though, according to the investigators, the "desire to equalize pay seemed 
here as sharp as desire to lighten work load." 

Such activity conventions, established by the work group, tended to 
reduce or even nullify the effectiveness of the wage-incentive system. The 
piece-rate system, instead of functioning as an incentive, served to stimu- 
late the development of an "informal organization" for the allocation of 
racks and for the collective control of rate of production apparently better 
designed to meet their needs or wants. Horsfall and Arensberg point out 
this informal organization might be called "restrictive," in that it was 
outside management rules. 

The discovery and description of the four teams and their informal or- 
ganization is not based solely upon interviews with workers or uncon- 
trolled observation. Findings of the type described above, and others dis- 
cussed below, follow from the use of an objective method of recording and 
measuring factual, nonverbal observations as a check on the evidence 
furnished by interviews. This technique, growing out of the earlier work 
of Chappie and his co workers (5,6), 9 resulted in a series of observation 
charts which provided (1) identification of the individuals observed, 
(2) discrimination of the roles of interaction (stimulus and response) 
between them, and (3) timing. Records of this kind were made for a 
total of 120 periods of 15 minutes each, 30 for each of the four teams. 
These were accumulated by an investigator suitably placed so that he 
could both observe the team and also hear most of the conversation which 
went on. However, the conversation was viewed as a supplement to the 
objective record of the interaction elements or "operations" as recorded 
on the chart. 

In Figure 19 is presented a facsimile reproduction of a completed ob- 
servation chart for a 1 5-minute period, together with a portion of the sup- 
plemental notes on the "content" of the interaction in terms of both 
actual activity and conversation. Although conversations were recorded in 
so far as possible, reliance is placed entirely on the objective recording 
of interactions to arrive at a description of the informal social organization 
of the group and of the "social" roles of its individual members. It is not 
feasible to present, in this volume, the detailed description of the formulae 
which would be necessary for the comprehension of the tables of findings 
presented by the investigators. According to Horsfall and Arensberg, the 
quantification of interaction verifies current hypotheses about inter- 
worker social relationships. The data, it is said, support the hypothesis 
that the clique-like groups represented by the four teams show a higher 

Further developments in the way of methods for studying social interactions as 
a process in small groups are found in the work of Bales (3) . 


Tuesday, August 8, 1938 










#100. R makes a comment 
to S. The assistant foreman 
passes down the aisle and 
pushes an empty rack out of 
his way toward mid-floor, 
whereupon T' calls to S to 
pet the rack back ; S gets it. 
R2 from down the room 
shouts, "Hey, the bar," 
whereupon T* mimics him, 
repeating, "Hey, the bar." 
(R asks T a question,) R3 
down the aisle, takes up 
R2's cry and T' then 
mimics him, R3 repeats, 
T' mimics again, this for 
several times. P is visibly 
annoyed by R3's shout- 
ing and finally shouts, "We 
got the box long ago," C 
leans over and shouts to S. 
The work is moving along 
at a good pace with R tak- 
ing the shoes from P as he 

does them,, no shoes having accumulated on R's bench. 

Figure 19. Graphic Record of Social Interactions in a Bottoming Room * 
(Note: In the diagram the letters at the top represent the members of the 
team, i.e. T = Trimmer, P = Pounder, etc. A wiggly line represents the dura- 
tion of an interaction extending over a minute, each horizontal line marking 
off 1 minute. The word "bar" refers to the object used in identifying types 
of shoes and in the allocation of racks.) 

rate of interaction among their members than between their members 
and outsiders. Of particular importance is the finding that teams with the 
most active internal social activity were not the most productive in terms 
of meeting output standards per worker or per team. The objective study 
of interactions led to the identification of a small number of workers as 
informal leaders. The teams of which the informal leaders of the room 
were members showed both a slight excess of internal activity over ex- 
ternal and a much higher activity of interaction with out-of-team, cross- 
team members of the room. In the phraseology of the foreman, the latter 
findings simply mean that "the teams that did the most talking and 'hors- 

* By permission, from A. B. Horsfall and C. M. Arensberg, 'Teamwork and pro- 
ductivity in a shoe factory," Human Org., 1949. 8, 13-26. 


ing around' did not get the most work done." One reason for this is that 
leadership, which necessarily involved "nonwork" activity, throws the in- 
formal leaders into enough extraclique, extrateam activity to make their 
teams score lower on "flow-of-work" activities in the immediate work 
group than do the teams who only respond in the room's own informal 
system of allocation and "controls/' 

Management's Role in the Development 
of Teamwork 

According to Horsfall and Arensberg, such findings "give pause to cur- 
rent enthusiasm about encouraging 'teamwork.' " Certainly such concern 
is justified if the "team" is defined as a clique-like group of workers which 
spontaneously "comes into being in order effectively to oppose demands 
from higher authority and work counter to the purposes set by manage- 
ment" (8, p. 47). 

The question may well be raised as to whether all such informal social 
organizations are or need necessarily be "subversive" of the purposes of 
the larger formal organization. Support for the view that this need not 
be the case is found in the conclusion by the Hawthorne investigators that 
in the original Relay Assembly Test Group behavior and sentiments were 
so organized as to promote the economic system of the plant ( 31 ) . Investi- 
gations conducted in the Army during World War II also indicate the 
distinct possibility that the informal group can play a positive role in bol- 
stering the formal or corporate organization (25). 


Studies in metal-working plants (9) and in the aircraft industry (17), 
under the direction of Mayo, have pointed to factors which need con- 
sideration in the striving for both spontaneous and organized teamwork 
productive of efficiency and morale. 

In these studies, conducted during World War II, attention was fo- 
cused on absenteeism and turnover as symptoms of unrest and insecurity, 
and as indices of high morale (17) or of involvement in achieving the 
objectives of the work group. As an outcome, in part, of these studies, 
Mayo and his associates have postulated the existence of three types of 
groups or work teams in industry, viz., natural, family, and organized. 

The natural and family groups are alike in that they "achieve their 
integration, not by direct action of management, but spontaneously and 
where management establishes around the workers a 'climate' of technical 
and operating aspects of organization such that groups can grow" (17, 
p. 23). 


The natural group is primarily a quite small group, varying from two 
or three to six or seven workers who are integrated into an informal social 
organization because of the "natural associative tendencies" of mankind, 
particularly if given any encouragement by management (24). 

The family group is a larger group, characterized chiefly by a central 
core of "regulars" who, "if they have prestige, determine, almost by in- 
advertence" or by example, the performance of its members. Such groups, 
according to Mayo and Lombard, actually develop only when a "natural" 
group can be held together long enough perhaps for a minimum of six 
months to a year to provide a core for the larger group. 

In contrast to such groups is the organized group which "comes into 
being only when someone in authority definitely works to create it. Man- 
agement has a direct relationship to the growth of the "organized" group. 
It differs from the other groups in that someone "with the respect and 
confidence of the workers and with the support of management . . . 
has set himself deliberately, with intelligence and skill, to achieve a group 
integrity of association, and to order the relations of his own integral 
group with other departments in the plant" (17, p. 23). 

A small department of 55 workers in an aircraft plant in Southern Cali- 
fornia furnishes an example of the "organized" group. This department 
had a good absenteeism record,, and the efficiency ran 25 per cent above 
the average of the plant. Approximately 90 per cent of the employees in 
the group were "regulars." A large proportion of these had no absences 
at all. "This situation," it is pointed out, "had not occurred by chance. 
The persons directly responsible were the senior assistant foreman and 
a lead hand" (24). Both of these were particularly concerned with the 
human aspect of their administration and showed this interest in their 
day-to-day activities in dealing with their men. Both believed that the 
achievement of group solidarity is of first importance in the plant and 
necessary for sustained production and, with the approval of the foreman, 
pursued activities designed to weld their men into an effective team work- 
ing toward the achievement of production and other management 


Consideration of studies discussed in this and other chapters indicates 
that management can help turn the attitudes and behavior of spontane- 
ously formed work groups into constructive channels. Management also 
has a positive responsibility for the direct organization of teamwork. In- 
vestigations conducted under the direction of Mayo, and others described 
in Chapter 7, suggest that keeping the work group small contributes to 


the organization of effective teamwork. Group (as contrasted with indi- 
vidual) incentive wage plans, may play a particular role in furthering the 
formation of teams characterized by constructive goals (17). According 
to Mayo and Lombard, the growth of teams oriented towards manage- 
ment's objectives can be facilitated by "building new work groups around 
common experiences and interests" and by the preservation of existing 
groups in the loan and transfer of workers (17, p. 3) . 


Underlying these approaches to the organization of teams is the view 
that positive effort in establishing "congenial groups of employees" can 
result in the improvement of output and morale (4) . Evidence supporting 
this view is found in a series of studies by Van Zelst. Findings from one 
of these studies, designed to explore the relation between job satisfaction 
and how well a worker is liked by his fellow worker, showed that the 
worker who rates high in "interpersonal desirability" was more satisfied 
with his job, had a greater feeling of job security, had a more favorable 
attitude toward the company, etc. (27). In another study, comparisons 
during a three-month experimental period showed that sociometrically ar- 
ranged work teams, i.e. groups combining individuals who are capable 
of harmonious interpersonal relationships, outperformed groups not so 
established and reported a higher degree of satisfaction with their jobs. 

The possibility of using sociometric methods for increasing production 
has been further explored by Van Zelst in a study of two groups of carpen- 
ters and bricklayers in the construction industry (28). The members of 
these groups, numbering 38 and 36 respectively, had worked together on 
the same building project for an average of at least five months and were 
well acquainted with each other's personality and skill. They were all 
union members, and none was a supervisory employee, or had less than 
seven years of experience in the trade. 

Prior to the experiment, each worker was asked to nominate three of 
his co-workers, in order of preference, as his choice of work partners. The 
workers were then regrouped on the basis of these choices, being first 
arranged into mutual choice teams of two. Because of the occasional need 
for larger work teams, patterns for the fusion of two teams into a single 
group on the basis of co-worker choices were also developed. Of the work- 
ers, 22 received their first choice as partners; 28 received their second 
choice, and 16 their third choice, 8 workers remained as isolates. At no 
time was it found necessary to cross skill levels and pair off a skilled worker 
with a relatively unskilled partner. 

The self -selected teams were used in subsequent work assignments with- 


out any change in the procedures for making duty assignments. 10 As a 
step in assessing the effectiveness of sociometric procedures in constitut- 
ing work teams, the production record of the sociometrically grouped 
teams for a period of 11 months was compared with engineers* estimates 
of cost (made for the purpose of submitting a contractor's bid) and with 
production records for 9 months prior to the restructuring of the work 
teams. Since the firm involved felt it undesirable to release figures on 
actual estimates and costs, labor and material cost indices (per unit row 
of 8 houses) are used in reporting the findings. 

Analysis of findings showed a labor cost index of 32.22 for the experi- 
mental period, as compared with 36.66 for the pre-experimental period 
of nine months and an engineers' estimate of 37.20. The material cost 
index proved to be 31.00 for the eleven months following the sociometric 
regrouping, as contrasted with 33.00 during the pre-experimental period 
and an engineers' estimate of 33.50. Differences between production 
records during the two periods proved to be statistically significant and 
amounted to a 5% savings in total production costs. In addition there 
was a significant drop in labor turnover during the second period. 

Preliminary comparisons showed that there had been no significant 
changes in the over-all skill levels of the total working group as a result 
of new hirings during the experiment. There were no changes in manage- 
ment practices in the course of the experiment and no differences between 
the experimental and pre-experimental periods in weather conditions. 
From these and other considerations the investigator concludes that the 
superior level of output during the experimental period is ''traceable only 
to the successful application of sociometric procedures and their effect 
in the work situation'' (28, p. 181 ) . 

In reporting the findings and conclusions of his investigation, Van Zelst 
notes that skill level was perhaps one of the standards used by the subjects 
in selecting work partners. It is therefore possible that the matching of 
skills in the sociometrically constituted work group may be an underlying 
factor in the improved productivity and reduced turnover. However, look- 
ing back at the findings of his series of studies, the investigator reaches 
the conclusion that "a careful employment of 'interpersonal relations' 
procedures increases the worker's sense of satisfaction and participation 
through an increase in his interest in and liking of his job, the removal 
of anxiety due to friction between work partners and the creation of a 
friendly, cooperative atmosphere" u (28, p. 184). 

10 Ten instances of change in the work-groups (4 voluntary and 6 for the incorpora- 
tion of isolates and of newly hired workers) were made during the first two-month 
period of the study. No disrupting incident occurred during the course of the experi- 

11 In this connection, further consideration might well be given to the fact that 



Van Zelst points out that the building trades with their "buddy-work- 
teams" are especially suited for a sociometric regrouping. Conversely, this 
technique may not be applicable or effective in other situations, particu- 
larly where changes in job duties, retraining, etc. would be involved. How- 
ever, even in such situations and universally throughout industry other 
tools are available to management for the effective organization of teams 
designed to serve both company and individual goals. 

Of particular significance, in this connection, is the accumulation of 
evidence which shows that the quality of supervision exercises a tremen- 
dous influence upon the behavior and attitudes of work groups, whether 
spontaneous in origin or organized under the direct influence of manage- 
ment. Thus, for example, preliminary interpretation of findings from the 
Relay Assembly Test Group experiment in the Hawthorne plant, dis- 
cussed in Chapter 10, led to the conclusion that the continued increase in 
production of this "informal social organization" resulted from a change 
in "mental attitude" toward supervision and toward the over-all environ- 
ment (21). 

A more recent review of the Relay Assembly Test Group experiment 
by Arensberg again calls attention to the significance of this situational 
change. "If," it is pointed out, "we simply rearrange the evidence in the 
order of the occurrence ... we see some new relationships. The five 
girls of the Relay Assembly Test Room developed their vaunted team- 
work after they were segregated by management as an experimental group, 
after their segregation brought them into much more nearly exclusive 
contact with one another, after they came into contact, consultation, and 
even veto-power relation with their supervisors, after, as they testified 
themselves, 'we have no boss/ All this makes a continuous change away 
from impersonal and authoritarian supervision a very sweeping shift in 
a basic relationship of modern industrial work in the direction of greater 
'participation/ . . . Teamwork did not come from the girls alone, what- 
ever their sentiments. It came after and thus perhaps from a sweeping 
change in supervision" (1, p. 340). 

Experiments and attitude surveys discussed elsewhere in this book 
support the conclusion that "the selection by management, through 

the subjects in the Relay Assembly Test Room in the Hawthorne series, where "social 
cohesion" most favorable to output was found, were selected on the basis of estab- 
lished friendliness to one another and because they were "willing, quick and co- 
operative" (22). By contrast, such criteria were not applied in the selection of the 
Bank Wiring Observation Group which, it is claimed, was characterized by a social 
organization unfavorable to output 


its acquaintance with its subordinates, of supervisors who can main- 
tain effective communication with their workers" is one way of pro- 
moting the growth of teams (17, p. 3). According to Dubin, the "organ- 
ized" group "is a product of the leadership exercised by the formally 
designated leader, the foreman" (8, p. 58). Such views find clear support 
in studies, described in earlier chapters, which show that the productivity 
and morale of the individual and of the group are functions of supervisory 
skill in tapping the worker's ego-motivation (14); in structuring the inter- 
actions of the members of his group (10); in providing opportunities for 
participation in setting goals and in other aspects of decision making, 
etc. Employee-attitude surveys described in Part IV of this volume reveal 
further the potent force of the quality of supervision upon the individual 
and upon the group.Studies by the Survey Research Center, described in 
Chapter 23, have shown that whether the worker is "union-minded" or 
"management-minded" is determined, in part, by the extent to which 
foremen involve workers in participation in decision-making (13). 

In general, consideration of research findings suggests that management 
must look to the foreman and supervisor for integrating workers into an 
"organized" group characterized by high levels of motivation and morale. 
This means, that management must give close consideration to the selec- 
tion of supervisory personnel and, most particularly, to the development 
of leadership qualities necessary for dealing effectively with human rela- 
tions on the job. It means also, as indicated elsewhere, 12 that management 
must furnish a "leadership climate" which supports the activities of 
"employee-oriented" supervisors capable of achieving a team integration 
characterized by "willingness to co-operate" in achieving both the eco- 
nomic and social objectives of industry (8, p. 58). 


1. C. M. Arensberg, "Behavior and organization: industrial studies" in J. H. Rohrer 
and M. Sherif (Eds.), Social Psychology at the Crossroads. Harper and Brothers, 
1951, 324-54. 

2. S. Barkin, "A trade unionist appraises management personnel philosophy," 
Harvard Bus. Rev., 1950, 28, 59-64. 

3. R. F. Bales, Interaction Process Analysis: A Method for the Study of Small 
Groups. Addison- Wesley Press, 1950. 

4. Y. Beaudry, 'The sociogram key to a better working team," Mill and Factory, 
(November) 1951, 49, 107-08. 

5. E. D. Chappie and C. S. Coon, Principles of Anthropology. Henry Holt & Co., 

12 See particularly Chapters 8, 17, and 23. 


6. E. D. Chappie (with C. M. Arensberg), "Measuring Human Relations/' Gen. 
Psychol. Monog., 1940, 22. 

7. P. F. Drucker, Concept of the Corporation. The John Day Co., 1946. 

8. R. Dubin, Human Relations in Administration. Prentice-Hall, Inc., 1951. 

9. J. B. Fox and J. F. Scott, Absenteeism: Management's Problem (Business Re- 
search Studies No. 29 ) . Harvard University, Graduate School of Business, Divi- 
sion of Research, 1943. 

10. N. Gekoski, "Predicting group productivity," Pers. Psychol., 1952, 5, 281-92. 

11. J. R. P. French, Jr. and A. Zander, "The group dynamics approach" in A. Korn- 
hauser (Ed.), Psychology of Labor-Management Relations. Proceedings of the 
1949 Meeting. Champaign, 111.: Industrial Relations Research Association, 1949. 

12. A. B. Horsfall and C. M. Arensberg, 'Teamwork and productivity in a shoe fac- 
tory/' Human Org, 1949, 8, 2, 13-26. 

13. E. Jacobson, "Foreman and steward, representatives of management and the 
union" in Human Relations Program of the Survey Research Center: First Three 
Years of Development. University of Michigan, Institute of Social Research, 
1950, 18-21. 

14. D. Katz, "Morale and motivation" in Current Trends in Industrial Psychology. 
University of Pittsburgh Press, 1949, 145-71. 

15. A. Kornhauser, "Analysis of 'class structure of contemporary American so- 
ciety* " in G. W. Hartmann and T. N. Newcomb (Eds.), Industrial Conflict. 
The Cordon Co., 1939, 199-264. 

16. K. Lewin, "Research center for group dynamics at MIT," Sociometry, 1945, 
8, 126-36. 

17. E. Mayo and G. F. F. Lombard, Team-work and Labor Turnover in the Aircraft 
Industry of Southern California, (Business Research Studies No. 32). Harvard 
University Graduate School of Business, Division of Research, 1944. 

18. E. Mayo, Human Problems of an Industrial Civilization. Harvard University, 
Division of Research, 1946. (First printed by The Macmillan Co., 1933). 

19. J. Mills, The Engineer in Society. D. Van Nostrand Co., 1946. 

20. J. L. Moreno, "Experimental sociometry and the experimental method in science" 
in Current Trends in Social Psychology. University of Pittsburgh Press, 1948, 
119-62 (See also J. L. Moreno, Who Shall Survive? (Mental and Nervous Disease 
Monograph Series No. 58). Washington: Mental and Nervous Disease Publish- 
ing Co,, 1934, and Foundations of Sociometry (Sociometry Monograph No. 4). 
New York: Beacon House, 1941. 

21. G. A. Pennock, "Industrial research at Hawthorne," Pers. J., 1930, 8, 296-313. 

22. F. J. Roethlisberger and W. J. Dickson, Management and the Worker. Harvard 
University Press, 1939. 

23. D. Roy, "Quota restrictions and gold-bricking in a machine shop," Amer. J. Soc. f 
1952, 57, 427-52. 

24. J. F. Scott, 'Teamwork," Modern Mgmt., 1946, 6, No. 3, 2-4; No. 4, 14-16. 

25. E. A. Shils, "Primary groups in the American Army" in R. K. Merton and P. F. 
Lazarsfeld (Eds.), Continuities in Social Research: Studies in the Scope and 
Method of "The American Soldier ." Free Press, 1950, Chapter 1. 

26. Triple Audit of Industrial Relations. University of Minnesota, 1951. 

27. R. H. Van Zelst, "Worker popularity and job satisfaction," Pers. Psychol., 1951, 

4, 405-12. 

28. R. H. Van Zelst, "Sociometrically selected work teams," Pers. Psychol., 1952, 5, 


29. I. R. Weschler, R. Tanncnbaum, and E. Talbot, A New Management Tool: 
the Multi-relational Sociometric Survey (Institute of Industrial Relations Reprint 
No. 25). University of California, 1952. 

30. T. N. Whitehead, Leadership in a Free Society. Harvard University Press, 1936. 

31. T. N. Whitehead, "Social motives in economic activities," Occup. Psychol., 1938, 
12, 271-90. 

Part 4 




Employee-Attitude Surveys: 
Scope, Methods, and 
Guiding Principles 

AMONG the major developments in industrial relations dur- 
ing the past quarter century is the growing awareness on the part of ex- 
ecutives that the ordinary channels of communication in industry fail to 
provide a clear or accurate picture of what is on the worker's mind. As a 
result, an increasing number of executives have turned to the employee- 
attitude survey to supplement and complement other channels of com- 
munication for "tapping employees' thinking" (39) . This is because they 
have found that such a survey can provide a systematic and comprehen- 
sive picture of workers' feelings, and reveal many specific relations which 
do not come to light through ordinary channels. "It is," as Kornhauser 
has suggested "like a thorough physical examination of the individual as 
contrasted with fragmentary information picked up casually concerning 
the state of his health" (61, p. 129). 



Evidence of accelerated use of employee-attitude surveys, especially 
since approximately the middle 1940's, is found in a series of investiga- 
tions conducted over a period of years by the National Industrial Con- 
ference Board. In a study of personnel activities made in 1939, involving 
2,700 companies, experience with attitude surveys was so limited that no 
question was even asked about them. In 1944, the Conference Board had 
difficulty in finding 50 companies which had conducted such surveys. By 



1947, a total of 245, or approximately seven per cent of 3,500 companies 
which were contacted reported that they had used one type of attitude 
survey or another. In a later study, published in 1951, the Conference 
Board found 223 companies which have had experience with attitude 
surveys (39). Of these, 111 companies which furnished detailed data 
reported that they had conducted a total of 201 employee-attitude sur- 
veys. However, only three of the companies involved had conducted sur- 
veys prior to 1940, and almost 86 per cent of the total number of surveys 
had been made subsequent to 1943. 


Consideration of the reasons given for conducting the surveys by the 
111 companies co-operating in the National Industrial Conference Board 
investigation shows that the vast majority undertook the surveys with 
definite goals in mind. More than a third "wanted to know how the peo- 
ple on the payroll regard one another, what they think about their jobs, 
their bosses, and about the company and the way it is run" (39, p. 9). 
Others were interested in detecting training needs, and in using survey re- 
sults for the evaluation of training. In some instances, there was a desire to 
discover what employees needed or wanted to know about the company, 
its operation, and other aspects of the business. In some cases, there were 
highly specific reasons, such as discovering the causes of high turnover, 
whether employees like the company magazine, etc. 

An interest in checking on morale conditions and finding ways to im- 
prove employee morale and relations within the company finds expression 
in reasons given for undertaking attitude surveys. Many companies ex- 
pressed a desire to learn about the minor troublesome situations so that 
measures could be taken to prevent their growing into major ones. In 
general, statements made by the companies show clearly an expectation 
that the attitude survey would provide management with a measure of 
its own success or failure in personnel matters and, at the same time, locate 
unsatisfactory feelings and sources of irritation requiring remedial ac- 
tion (61). 


It is apparent that employee-attitude surveys are viewed by manage- 
ment as a practical tool which can be used to help uncover and solve plant 
personnel problems. Attitude-survey techniques are of additional interest 
to the psychologist as a tool for obtaining deeper insight into the nature 
and interrelationships of basic motives and allied psychological processes. 
As used in this connection, particularly as adjuncts to field experimenta- 
tion, employee-attitude surveys are intended to yield broad principles 


bearing upon the problems of productivity, satisfaction, and morale, and 
generalizations which have significance in terms of motivational theory, 
rather than a series of facts which are of interest only in relation to the 
immediate needs of a particular company. 

As Katz has pointed out, surveys conducted for the latter purpose may 
have as their by-product some contribution to the psychology of the 
worker (57). However, such surveys may contribute little to a science 
either of industrial psychology or of industrial relations because, as is 
frequently the case, they do not employ a research design adapted to a 
systematic analysis of the variables involved and to establishing the rela- 
tionships especially the causal relationships among these variables. An- 
other deficiency of many investigations, from the viewpoint of the broader 
implications of research in the social sciences, is the failure to design and 
conduct the surveys within a context determined by sound theory. 

Studies described in Part III, such as those conducted by the University 
of Michigan Survey Research Center, illustrate the use of employee- 
attitude surveys as a methodological tool in such a context. Of special 
interest in such studies, as compared with the typical management-spon- 
sored survey, is the increasing effort to use survey techniques to throw 
further light upon the group process "the motivation for, as well as the di- 
rection of the behavior stemming from the group, and taking place within 
an institutional context" (71). In general, the survey method, most 
frequently used as a practical tool, has wide applicability as a research tool, 
especially when used in a "well-designed research plan" (58). 

Employee-Attitude-Survey Methods 

The techniques employed and the theoretical and practical problems 
involved in surveying attitudes and opinions have been discussed by 
Thurstone and Chave (109), Likert (68), Doob (35), Cantril (18), 
Parten (84), Stouffer et al. (107), Blankenship (8, 9), Day (30), Mc- 
Nemar (72), Guttman (44, 45), Churchman, Ackoff and Wax (25), 
Jahoda, Deutsch, and Cook (52), and others (17, 21, 26, 36, 40, 99, 117). 
A practical guide for use in preparing and administering an employee- 
attitude survey, and in analyzing results, is found in a National In- 
dustrial Conference Board report on company experiences with such 
surveys (39). Critical issues in the use of survey techniques for practical 
and research purposes have been summarized by the author (113) and 
also by Kornhauser (61), Deming (31), Katz (57), Ackoff and Pritzker 
(1), and Worthy (121). 

It is not within the scope of this volume to present a detailed discussion 
of how to conduct an employee-attitude survey. However, it seems desira- 
ble, as a preliminary to the discussion of survey findings, to outline (1) 


the techniques used in ascertaining employee attitudes and (2) considera- 
tions -which should be borne in mind in evaluating employee attitude 
survey findings. 


Procedures most frequently used (either separately or in combination) 
in ascertaining employee attitudes can be summarized, following Korn- 
hauser (61), as including 

1. Printed Questionnaires: In most instances employee-attitude sur- 
veys have made use of a series of questions or items presented in writing 
and requiring written responses by the participants. These, as illustrated 
in sample questions on page 225, generally include "yes-no," "true-false" 
items, "check-lists" and also, in many instances, provide opportunity for 
written-in comments related to a particular question, or to items not 
covered in the form, or both. A special form of the printed questionnaire 
is found in the attitude scale, constructed through the use of specialized 
calibration techniques to provide a list of items of varying degrees of 
favorableness regarding the matter in question. A portion of such a scale 
is reproduced on page 273, in connection with the discussion of the sur- 
vey reported by Uhrbrock. 

2. Interviewing Techniques: Many employee-attitude surveys have in- 
volved the use of interviewing techniques. In an increasing number of 
cases, even when the main body of data are obtained by means of ques- 
tionnaires, interviews with a sample of employees are used both to check 
on questionnaire findings and to elicit additional information (in depth) 
with respect to attitudes. 

Interviews are of both the guided (directive) and unguided (nondi- 
rective) type. Various degrees of patterning have been employed in 
guided interviews, ranging from rigid adherence to formalized questions 
to a minimum degree of control. In some cases, direction extends to the 
point of patterning the interview around a series of formal interview ques- 
tions with simple choice responses, similar to those included in printed 
questionnaires, which are answered orally instead of in writing. Generally, 
a moderately formal interview schedule is followed and, even in such 
instances, the participant may be encouraged both directly and through 
follow-up questions to volunteer his own ideas and to express his feelings 
on whatever matters in whatever way he prefers. 

In contrast with a complete or modified directive approach is the un- 
guided interview, in which the employee is encouraged to talk about what- 
ever is on his mind. Here, the function of the interviewer is to listen and 
to encourage the participant to talk, but without asking specific questions. 


Sample Attitude Survey Questionnaire Items: Caterpillar Tractor Company (from 
Survey Research Center Study, after National Industrial Conference Board, 1951). 


16. How much are you bothered on your 
job by each of the things listed be- 
low? (Check each item) 


Both- Both- both- 
ered ered ered 
a lot some at all 

Noise l D D D 

Dirt, oil, or grease 2 Q Q D 

Lighting 3 Q Q Q 

Dust or fumes 4 . . . D D D 
Heat or cold 5 . . . . Q Q D 
Condition of wash 

room n D D 


17. What shift do you work on? (Check 

D 1st shift* 
D 2nd shift * 
D 3rd shift * 

18. How do you like working on this 
shift? (Check one) 
n I like this shift best * 

It makes no difference to me 2 

I would rather work some other 

shift 8 

I would much rather work some 




other shift 

20. Do you usually buy your lunch in the 
cafeteria? (Check one) 

D Yes 1 D No* 

(Please answer questions 21 and 22 
as well as you can whether you eat in 
the cafeteria or not.) 

21. Company cafeterias are run on a non- 
profit basis. Considering this, how do 
you feel about meal prices in the caf- 
eterias? (Check one) 

Q Prices are much too high x 

D Prices are somewhat too high 2 

n Prices are about right s 

D Prices are fairly low * 

D Prices are very low * 

22. What do you think of the food in the 
cafeteria? (Check one) 

n The food is much better than 

you can get other places for the 

same money x 
n The food is somewhat better 

than you can get other places for 

the same money 2 
n The food is about the same as 

you can get other places for the 

same money 8 
D The food is not quite as good as 

you can get other places for the 

same money 4 
n The food is not nearly as good as 

you can get other places for the 

same money 5 

This procedure is illustrated by the employee interviews in the Hawthorne 
studies, referred to on pages 254-258. 

The methods described above, with the exception of the free-response 
and highly nondirective interview, are essentially structured in character. 
Moreover, all of them are inherently overt, or "nondisguised" methods 
of attitude surveying. In the same general category, at least in the sense 
of overtness, are the sociometric techniques, developed by Moreno (24, 
42, 70a, 74, 75, 76) which have been applied in studying "the evolution 
and organization of groups" (75) and analyzing sentiments and attitudes 
relating to social groupings and social interactions. The basic approach 
here is to ask each member of the group to express an attitude toward 


his fellow members, although there are many variants of this procedure, 
including graphic plots of the relationships within the group * (53, 53a) . 
Likewise, the intent of the tetter- writing technique, used in the General 
Motors Corporation's My Job Contest (64), described in Chapter 13, 
is easily recognized, and the technique also appears to be quite highly 
structured, in spite of the apparent freedom allowed to respondents. 


Such methods have been supplemented by a series of "disguised-non- 
structured" methods for the measurement of attitudes. Underlying such 
indirect approaches in attitude surveying is the viewpoint that the overt, 
formal and rigid structuring of conventional survey techniques tends to 
obscure true employee attitudes (38). As Weschler has pointed out, the 
purpose of indirect methods of attitude measurement is to "get at those 
'deeper lever attitudes which a person may be interiorizing and unwilling 
to reveal. These indirect devices" he adds, "conceal from the individual 
the intent of the measurement and allow him to produce responses which 
would not be freely forthcoming if he were fearful of becoming personally 
involved" (116, p. 133). 

Indirect procedures for the measurement of attitudes have been used 
to a very limited extent in industrial surveys, and will therefore not be 
described in any detail in this volume. The methods employed have been 
described in an extensive review by Campbell (17) and, more recently, 
by Weschler and Bernberg ( 117) . Among "disguised-nonstructured meth- 
ods" are protective techniques (6, 41), which include the use of picture 
interpretation (78), as illustrated in research of interest to industry by 
Proshansky (86), Haire and Gottsdanker (47), Ehle (37), and Sanford 
(94, 95, 96) . Among projective techniques, play and dramatic materials, 
as employed for example in a study by Dubin ( 34) , have been described as 
offering interesting possibilities as indirect means of eliciting social atti- 
tudes (117). 

Projective techniques include a series of devices involving the use of 
written and verbal materials in the measurement of attitudes. The word- 
association, the sentence-completion, the short-story completion tests 
are among such devices, as are also "a series of projective methods which 
allow the subject to call upon his literary imagination and to project his 
feelings and attitudes into the completion of arguments, the invention 
of story plots, or the writing of autobiographical materials" (117, p. 221 ) . 
An interesting example of the use of the argument completion technique 
(77) is found in a study by Jones and his coworkers of the social system in 
a factory town (55). 

i Sec pages 200-205; 207-208; 213-214. 


Projective methods for testing attitudes both disguise the purposes of 
the test and are nonstructured in character. There are, in addition, a 
number of techniques for measuring attitudes which disguise the purpose 
of the test but make use of relatively structured situations. In general, 
these involve "an attempt to diagnose attitudes from systematic bias in 
the performance of an objective task" (17, p. 19). Such an approach in 
attitude testing is found in the "error-choice" method, originally de- 
veloped by Hammond (46). This undertakes "to relate the permanence 
of error in perception to the measurement of attitudes by provoking the 
subject to sift pseudo-facts from memory and measuring the 'direction' of 
the error" (117, p. 222). As indicated on page 336, Weschler (115) 
has adapted this technique to the measurement of attitudes toward labor 
and management. More recently, Bernberg (7) has used a further ad- 
aptation of this technique, designated as the "direction of perception" 
method, in the measurement of the morale of employees in an aircraft 
manufacturing plant. 

Developments in this area of attitude surveying include a number of 
tests employing attitudinal bias in perception and memory, based on 
systematic studies of such bias in "objective" perceptual and learning 
tasks (2, 67, 100, 98, 51). A large number of methods for the disguised 
measurement of attitudes by means of "objective tests" as well as by 
tests involving perception and memory distortion have been developed 
and tried out by Cattell and his associates (20, 22, 23) . Among disguised, 
structured and nonvoluntary methods for assessments listed by Campbell 
in his extensive review of indirect methods are estimations of group opin- 
ions and social norms, and tests involving ability to judge character. 

It is apparent that a variety of indirect methods have been explored 
in the effort to develop techniques which will avoid dependence upon 
"voluntary self-description" and tap the "deeper levels" of attitudes. 
Nevertheless, indirect methods have not yet attained the applicability of 
the conventional methods, or (in spite of the many inadequacies of the 
latter) achieved the same status in the methodological terms of reliability, 
validity (17) and, as McNemar would require, unidimensionality (72). 
As Weschler and Bernberg have pointed out, "Indirect methods of at- 
titude measurement have established themselves as useful and provoca- 
tive research tools, but their future status and development depend upon 
the type of research which will stress methodology rather than originality, 
and scientific technique rather than ingenuity of design" (117, p. 225). 

Problems in the Evaluation of Survey Data 

Surveys using the techniques described above, and occasionally others, 
have yielded a vast amount of data concerning employee attitudes. Un- 


fortunately, the evaluation of such data is a much more difficult matter 
than their collection. As will appear in later chapters, there are many in- 
stances of factual discrepancies among the findings of different surveys. 
There are also differences in the meanings attached to such findings which 
reflect, to some extent, varying frames of reference in the interpretation 
of data. 

In general, in dealing with individual studies, an effort has been made 
to identify the sources of discrepancies and to discuss problems of inter- 
pretation. However, it seems desirable to outline systematically and in 
some detail, the factors which should be borne in mind in evaluating 
survey findings. 2 This is intended primarily as a guide for those who are 
not survey specialists, but who have a professional or scholastic interest 
in the methods and problems of surveying attitudes. There is also the 
possibility that this section may be of service in plotting, for the execu- 
tive, the shoals and reefs in the sea of apparent contradictions in which 
he at times finds himself floundering when he undertakes to use survey 
findings in steering a course toward the goals of high productivity, satis- 
fied employees, and harmonious labor relations. 


Discrepancies in survey findings (which necessarily affect conclusions 
drawn from the surveys) arise, in part, from the varieties of questionnaires 
and items used in such study (10, 114) . Surveys which, like that described 
on pages 258 ff., call only for ranking a list of "morale factors" may 
well produce different results from those which make use of multiple- 
choice items calling for an expression of feelings with respect to specific 
plant conditions, policies or practices. 

This situation is further complicated when there are imperfections in 
the design of questionnaires. Among such imperfections, as listed by Dem- 
ing (31), are lack of clarity in definition and ambiguities in the phrasing 
of questions (81, 85). Questionnaires may produce errors in surveys be- 
cause various meanings are attached to the same word by different groups 
of people, and because of replies which are liable to misinterpretation. 
The omission of questions which would be illuminating constitutes an- 
other source of difficulty. Involved also are emotionally toned words, lead- 
ing questions, and the influence of the rigid patterning of responses (29) . 

An illustration of the influence of question phrasing is found in a mar- 
ket survey conducted on behalf of a national publication. One of the ques- 
tions asked was: "Do you subscribe to ?" Internal incon- 
sistencies in the findings raised some doubts concerning the meaning 

2 In the development of this section the author has drawn particularly from articles 
by Kornhauser (61), Katz (57), and Deming (31). 


attached to this question by respondents. The use of another sample, in 
the same geographical area, showed clearly that to some people the sen- 
tence "Do you subscribe to ?" meant "Do you accept the 

editorial policies of ?" and not, as intended, "Do you take 

; regularly through the mail? 9 ' 

The possible biasing effect of the question can also be illustrated from 
an employee-attitude survey, reported by Factory (118) . One of the ques- 
tions used read as follows: "Here's another idea. There are probably some 
men in your plant who loaf a little bit. Now suppose nobody loafed and 
everybody turned out just as much work as he reasonably could would 
this help or hurt the workers?" It is apparent that the answer given to this 
question may well depend on the respondent's interpretation of the 
phrase u loaf a little bit" and that the final conclusion that "AH types of 
employees frown, however, on loafing" may be suspect because of the bias- 
ing effect of the term "little bit/' 

Survey specialists are naturally aware of these hazards in questionnaire 
construction. Nevertheless, as will appear from the discussion of indi- 
vidual studies, imperfections do creep in and are at times overlooked in 
the interpretation of findings. Moreover, a failure to formulate the theo- 
retical context of the investigation, or unfamiliarity with the details of a 
situation, or even technical difficulties (such as arise, for example, in deal- 
ing with the effects of the "group" on individual attitudes) may result in 
the omission of items which are crucial to an adequate portrayal of senti- 
ments, feelings and attitudes. 


The problem of questionnaire items arises not only when printed forms 
are completed in writing, but also in interview surveys when questions 
are formulated for use in highly patterned form (105). In less patterned 
interviews, variation in findings (both within a survey and between sur- 
veys) can result from the bias of the interviewer as derived from both the 
personal characteristics and beliefs of different interviewers (11, 40a, 56, 

This situation is well illustrated in the classic investigation by Rice 
(89), who analyzed the records of interviews by 12 investigators of 
2,000 homeless men, applicants for lodging, and discovered evidence of 
a contagious transfer of each investigator's individual bias to the appli- 
cants with whom he talked, and a corresponding distortion in their re- 
plies to scheduled questions. One of these interviewers was an ardent 
prohibitionist. He found the downfall of 62 per cent of the applicants 
to be due chiefly to liquor, while only 7 per cent had been seriously affected 
by industrial conditions. Another interviewer, a Socialist, found that only 


22 per cent owed their plight primarily to liquor, and that 39 per cent 
had been affected chiefly by industrial causes. Moreover, the prohibi- 
tionist reported that 34 per cent of the applicants themselves gave liquor 
as the cause, and 42.5 per cent industrial conditions, but the Socialist 
quoted only 11 per cent as blaming liquor and 60 per cent as blaming 
industrial conditions. Each interviewer was a trained and conscientious 
investigator, but his personal bias affected his interpretation and report, 
and even modified the statements made to him by the men he questioned. 

Such bias has been described by Woodworth (120) as the "preconcep- 
tion controlling the interpretation" of facts. Involved here is the whole 
process of "social perception" the tendency to "read meaning or sig- 
nificance into people or situations with which we have had no prior ex- 
perience in terms of the assumptions we have built up that are apparently 
most relevant in terms of our own purposes" * (19, p. 140). 

As Deming has pointed out, variations attributable to the interviewer 
can arise from the political, social, and other beliefs of the interviewer, 
from his educational and economic status, and from his environmental 
background. Variations attributable to the interviewer can also arise from 
differences in the degrees of rapport established by interviewers. The 
personality and approach of the interviewer may affect the moods and 
behavior of respondents, i.e. "the interviewer may make the respondent 
gay or despairing, garrulous or clammish . . . cause respondents to take 
sides with them, some against them" (31, p. 363). 

Lack of knowledge or understanding of the content and purpose of 
the survey can constitute a source of bias and variability arising from the 
interviewer. In the absence of such understanding, the interviewer may 
not be able properly to record the respondent's statement or evaluate 
the situation. Conversely, highly trained interviewers, thoroughly indoc- 
trinated with respect to the purpose of the survey, may unintentionally 
be swayed in recording and evaluating responses by the interests to be 
served by the investigation. A number of studies reported in this volume, 
such as the Hawthorne studies, need close examination in terms of the 
latter possibility. Involved here, in part, is an effect of the "bias of aus- 
pices," discussed on pages 232-235. 


One difficulty in the evaluation especially the comparative evalua- 
tion of survey data arises from limitations in the samplings of people 
in such studies (32, 72, 84). This is generally no problem in surveys con- 
ducted in individual industrial plants which include all employees who 
complete questionnaires on company time, under the supervision of an 

See pages 93-96. 


employee committee or of an outside organization. Serious questions, 
nevertheless, can arise when mailed "ballots" are used; when the co- 
operation of employees is sought on an "invitational" basis; where the ad- 
ministrative conditions are such as to lead a substantial portion of the 
group to refrain from completing the forms; and under other condi- 
tions (15, 88, 101, 103, 108). Furthermore, there appear to be at least a 
few instances in which selective techniques have been employed to yield 
far from a representative sampling of employees in the industrial plant. 

The problem becomes more complex in studies calling for a repre- 
sentative sampling of industrial situations. It is entirely possible, for ex- 
ample, that plants which agree to co-operate in large-scale surveys, such 
as those conducted by the National Industrial Conference Board, repre- 
sent a different "universe" than those who decline to co-operate. Similarly, 
union plants may well yield different findings than do non-union plants. 
In most instances, the latter factor has been given close consideration in 
studies involving nation-wide samplings of workers and supervisory per- 
sonnel, such as those conducted by the Opinion Research Corporation 
and under the auspices of Factory, Fortune and similar publications. 

Such nation-wide surveys, nevertheless, raise additional problems in 
the way of the possibility of biased samplings. For example, as Deming 
points out, an interviewer who has been told to talk to ten people in a 
particular socio-economic or other group may choose to talk to ten peo- 
ple close to his home who will respond to his questions. As a result, the 
interviewer, in filling his quota, introduces a "respondent bias" that may 
take on considerable magnitude (31). Investigators are thoroughly aware 
of this danger but this, nevertheless, represents one of the considerations 
which must be kept in mind in evaluating samplings in employee-atti- 
tude surveys (73, 12). 


Closely related to the general problem of sampling discussed above 
is the selective bias of situations. This occurs in employee-attitude sur- 
veys from the fact that "by the nature of the case, most of the inquiries 
have been conducted in companies with enlightened and progressive 
managements, and ones where conditions of morale are reasonably satis- 
factory. Managements which are sitting on a powder keg," writes Korn- 
hauser, "do not conduct employee attitude studies" (61). 

Neither a union leader nor management, according to Katz, welcomes 
surveys when they are deadlocked in a struggle (57). As a result, much 
valuable data concerning unfavorable attitudes during particularly critical 
situations may be missing from available reports. Conversely, some com- 
panies may inadvertently or otherwise conduct a survey immediately fol- 


lowing a wage increase or under otherwise favorable circumstances which 
distort the findings in terms of the long-range situation. Such hazards 
are most frequently avoided in nation-wide surveys, which include sam- 
plings of employees from many plants, but this potential must neverthe- 
less be seriously considered in evaluating attitude-survey findings. 


As suggested above, there are influences derived from the period in 
which a survey is conducted. A survey of household purchases in the weeks 
before Christmas cannot provide a true assessment of the year-around 
market for "luxury" or possibly other products, although indicating trends 
for the holiday season. A survey of employee attitudes during the height 
of a depression or the peak of a war emergency situation can provide data 
applicable to the period under consideration and for comparisons with 
other periods. However, care must be exercised in formulating broad 
generalizations without confirmatory evidence accumulated during pe- 
riods varying in social, economic, and other conditions. 

Consideration of the factors discussed above underlines the need for 
careful description by investigators of the situation within which trends 
or relationships have been observed. It also indicates the continuing need 
for the multiplication of studies in situations characterized by varying 
conditions, as a basis for extending the scope and degree of generalization 
of conclusions. 


Another problem which arises in attitude surveying involves a possible 
bias of the auspices. For example, according to Deming (31), it is com- 
monly supposed that replies concerning such objective facts as income 
and work status are different, on the whole, when elicited by a relief 
agency than when information is sought by a government agency, such 
as the Census Bureau. Recognition of the possible import of bias of aus- 
pices is found, to some extent, in the frequency with which industry 
makes use of outside agencies in conducting employee-attitude surveys, 
although the primary emphasis here is generally that of giving employees 
assurance of anonymity of responses and of complete objectivity in analyz- 
ing the data. However, the fact that the vast majority of employee atti- 
tude surveys are management-sponsored has led some union leaders and 
some social scientists to question both the findings and conclusions from 
such surveys. 

Differences in evaluation associated with sponsorship are strikingly 
illustrated in comparisons of executives' and union leaders' assessments 


of the frankness and honesty of opinion expressed by employees in atti- 
tude surveys. According to a National Industrial Conference Board sur- 
vey, as reported in 1951, four out of five of the company co-operators felt 
that employees gave frank and honest opinions with few or no exceptions. 
By contrast, nearly all of the 43 union leaders who were questioned "state 
or imply that what employees say in a survey will doubtless be affected 
by the answer to the question 'Who's behind this survey?' " (39, p. 45) . 
Furthermore, about 85 per cent expressed the opinion that surveys should 
be conducted by "the union and the company/' 

The bias of auspices, it has been suggested, may be reflected both in 
the choice and phrasing of questionnaire items or interview queries. For 
example, in the union-sponsored survey (91 ) described on pages 346-352, 
workers were asked: "Since you are a member of a union, do you feel that 
your boss can't push you around as much as he might want to?" The 
implication that the boss wants to push workers around, may well pro- 
duce a "one-sided focussing" of attention which rather than a critical 
weighing of pros and cons of the issue will determine the replies (29). 
This issue, particularly as it affects the measurement of attitudes toward 
labor, has been considered by Kornhauser (59) and also, in more general 
terms, by Crespi (29). The latter draws upon Kornhauser for an illustra- 
tion pertinent to this discussion. 

"Suppose a ballot sought to ascertain people's attitudes toward the 
labor movement, with most of the questions being similar to the fol- 

"Westbrook Pegler, the newspaper writer, claims that many labor 
union leaders are racketeers. Do you agree or disagree with him?" 

"To make more jobs, some unions require employers to hire more 
persons than are actually needed to do the work. Would you favor or 
oppose having a law passed which would stop this practice?" 

"Do you think that soldiers, when they come home, should have to 
join a union in order to get or hold a job?" ( 59 ) . 

With a preponderance of questions of this type, according to Crespi, 
the respondent is forced to preoccupy himself with the demerits of or- 
ganized labor. "This dwelling upon negative and disapproved features," 
he adds, "can hardly fail to influence the respondents' attitudes in a nega- 
tive direction" (29, p. 105). This situation can be remedied, according 
to Kornhauser, by consistently maintaining an "equitable balance" in 
surveys, e.g., by matching any question which refers to the faults of labor 
with a question bearing upon the good points of labor. This position has 
been questioned by Link and Freiberg (69) . It has also been suggested that 
this problem might be solved by the more frequent use of neutral ques- 


tions. Involved in this issue, according to Crespi, is not alone the objec- 
tive of obtaining measures of true attitudes, but of refraining from using 
a survey as a propaganda device in itself. 

Differences in theoretical context or in frames of reference, reflecting 
in some instances the auspices under which a study is designed, can seri- 
ously affect the findings and generally influence the interpretation of re- 
sults. This possibility is well illustrated in surveys of employee attitudes 
toward the Taft-Hartley Law, discussed in Chapter 19. This problem will 
be referred to again below under the discussion of the interpretation of 
survey findings. 

One other issue remains to be mentioned under this discussion of 
the bias of auspices. Reference has been made to the fact that, except 
for some government or foundation-supported research, employee-atti- 
tude surveys in industrial plants have generally been sponsored by man- 
agement and used for management purposes. According to some ob- 
servers, as represented by McConnell, "this is unfortunate because the 
problems to which attitude surveys have been applied have been dictated 
by management needs" (71, p. 224) . According to Barkin, Director of Re- 
search, Textile Workers Union of America, CIO, "Whatever value it 
might have in simple areas of opinion, the self-reporting technique on 
attitudes and opinions in the field of industrial relations, as currently 
used, is dangerous and reinforces anti-union employer philosophies" 
(5, p. 80). 

Associated with queries concerning the management sponsorship of 
employee-attitude surveys is the tacit or stated assumption that union 
sponsorship, or the assignment of responsibility for conducting surveys 
to socially-minded scientists, would somehow or other bring about a 
more exact determination and satisfaction of employee wants and needs 
(60). There is no question that unions, if they were willing, could use 
survey techniques to advantage in determining the attitudes of their mem- 
bers more accurately than is frequently done by union leaders. Certainly, 
problems of organization, leadership, and morale in the unions provide 
a fertile field for the use of survey techniques in the interest of more 
effective service to members, as has been demonstrated in the study of 
the attitudes of union members by Rose (91). Furthermore, benefits 
might be achieved by joint, i.e. management-union sponsorship of sur- 
veys in unionized plants (41a, p. 11 ) . There can be no disagreement with 
the need for the design and administration of surveys by independent 
research groups, in a social frame of reference and a theoretical context 
which will enhance the usefulness of the social sciences in dealing with 
the problems of industrial strife as a broad social issue. 

The fact also remains that employee-attitude surveys sponsored and 


supported by management are serving a broad social purpose, as well as 
the interests of the individual plant and of its workers. In fact, the bias 
of auspices in this case may well be viewed as simply another indication 
of the growing effort on the part of management to arrive at a better 
understanding of employees' wants and needs, in part because of growing 
conviction that the satisfaction of such needs is essential to the accom- 
plishment of broad economic and social objectives as well as the immedi- 
ate aims of the business enterprise. 

In other words, the position of this author is that the mere fact that 
a survey is supported by management and is frankly intended to serve 
"management's needs" does not in itself constitute a reason for ques- 
tioning the outcome, either in terms of benefits to the workers in the 
plant, or in terms of consideration of the data together with those from 
other surveys in arriving at generalizations on motivation and morale 
in industry. There seems reason for questioning the view, held by some, 
that only the social scientist, working outside of the realm of problems 
of immediate interest to management and other groups, "has a job to do 
for democratic thought and practice" (66). Nevertheless, the principle 
of inquiring into the auspices whether they be management, union, or 
presumably impartial socially minded investigators represents one step 
in a comprehensive and sound approach to the evaluation of survey find- 


Differences and deficiencies in criteria and methods used in tabulating 
and analyzing data introduce variations which require serious considera- 
tion in the evaluation of survey findings. Such variations arise, in part, 
from the failure to perceive which tabulations and methods of analysis 
will be most significant (72). 

As suggested below, what will be available for tabulation and analysis 
depends, in part, upon the research design and the theoretical context of 
the investigation. In fact, as Deming points out, the tabulation program 
should precede the construction of questionnaires, etc., although this 
order is seldom followed. However, even where the basic data are relatively 
adequate, tabulations and methods of analysis may not be appropriate 
for revealing the complete picture. For example, a difference may be 
found in the level of morale of two work groups engaged in the same 
kind of work in the same plant, but under two different supervisors. It 
might immediately be concluded that the difference in supervisors ac- 
counts for the observed variation in morale level. The two groups may, 
however, differ also in age, amount of education, length of service, and 
in other respects. The investigator may be led astray in formulating con- 


elusions unless tabulations of age and similar factual data are available 
and are considered along with the analysis of questionnaire or interview 
data. Similarly, comparisons of the morale of short-service and long-service 
employees, such as are considered in Chapter 14, may be suspect when 
there is no assurance that older employees with longer years of service 
are holding the same kinds of jobs as the younger, less experienced em- 
ployees with whom they are compared. 

Involved here is the problem of controlling or eliminating the effects 
of "extraneous" variables which plagues the experimenter as well as the 
attitude-survey specialist. In this connection, Kornhauser has called at- 
tention to the prevalent use of simple correlational techniques in situa- 
tions where the techniques of partial correlation are required to reveal 
the exact nature and extent of relationships (61 ). Equally basic is a tend- 
ency to rely on counting procedures, e.g., on determining the percentage 
of workers who like the retirement plan, who have confidence in top 
management, etc., in instances where the full needs of the situation can 
be made clear only by an analysis of relationships ( 57) . 

An extension of the analysis to include treatment not warranted by 
the available data represents another source of error in the evaluation 
of attitude survey findings. Instances of this are noted in the discussion 
of individual studies in later chapters. Furthermore, the identification of 
factors related to morale presents special problems in the analysis of sur- 
vey results. These are discussed in Chapter 14. 


In attitude surveying, as in other activities involving measurement, it 
is necessary to determine the reliability of the measures. Reliability bears 
upon the question of whether the attitudes or morale score obtained by 
the individual would be the same if the same questionnaire or interview 
were repeated under the same set of known conditions (43, p. 411). It 
refers essentially to the consistency of the measure and may be defined in 
terms of "the degree of error involved in assigning an individual to a class 
or in establishing his rank-order position" (72, p. 294) . The import of re- 
liability in opinion and attitude surveying has been well stated by Mc- 
Nemar in the question: "If an opinion or attitude is such a momentary 
thing that individual reliability or consistency is lacking, how can one ex- 
pect the opinion in question to correlate with stable sociological or psy- 
chological characteristics of the individual?" 4 (72, p. 295). 

Reliability: Reliability is ordinarily expressed in the form of the con- 
ventional coefficient of reliability obtained by correlating alternate forms 
or sets of measures derived from a single scale or from interview data. 

4 Not italicized in the original. 


The use of test-retest correlation even more than in the case of test con- 
struction presents difficulties because of the possible effects of inter- 
vening experience upon attitudes. 5 Involved is a basic problem which 
cannot be discussed in detail here, although the discussion of methods 
for changing attitudes, in Chapter 22, has some bearing upon this "di- 

In his extensive review of opinion-attitude methodology, McNemar 
has noted many instances of scales with low reliability (72). This situa- 
tion appears to exist also in employee-attitude surveying. Examination of 
studies in this field further reveals many instances in which no data on 
the reliability of obtained measures are given. Such omissions introduce 
difficulties in the interpretation of findings. However, there is an increas- 
ing recognition of the need for adequate reliability and also, as appears 
in later chapters, many cases in which acceptable reliability of employee- 
attitude measures and morale scores are reported in the literature. 

Validity in Predicting Behavior: Many complex problems exist in as- 
sessing the validity of attitude measures. In general, it is necessary to dis- 
tinguish between validity in terms of prediction from survey data, and 
validity in terms of interpretation of attitude survey data (83) . 

The object in aptitude-, interest-, and similar test-construction programs 
is that of developing a measure from which some other kind of behavior, 
such as academic success, job performance, turnover, etc., can be pre- 
dicted. Similarly, in opinion and attitude surveying, the validity of a meas- 
ure can be viewed in terms of the extent to which the measure will pre- 
dict some kind of behavior on the part of the individual (70). So, for 
example, opinion polls have been conducted on the assumption that, 
from the measurement of attitudes toward candidates or issues, predic- 
tions can be made as to voting behavior. Similarly, in industry, the prob- 
lem of validity has been approached from the viewpoint of the usefulness 
of attitude measures and morale scores in predicting production, turn- 
over, occurrence of conflict situations, etc. 

The tendency in earlier days of employee-attitude surveying was to 
take it for granted that behavior could, in fact, be predicted from the 
findings of employee-attitude surveys. Here, as in public opinion survey- 
ing, there was rather general acceptance of the categorical point of view 
that behavior was closely related to observed attitudes. Behind this view 
is, as indicated on pages 73-75, the theoretical conception of attitudes 
as "integrations mediating between the fundamental psychological proc- 
esses and actions" (62, p. 31). However, a one-to-one relationship be- 
tween attitude and action has failed to appear in a number of studies, 

6 The possible effects of cyclical changes in feeling-tone, as observed in experiments 
by Johnson (54) and Hersey (48, 49), also need consideration in this connection. 


as exemplified in those conducted by Corey (28), La Piere (63), and 
Bray (12a). In more recent years, particularly as a result of a number of 
outstanding failures of polls in predicting presidential elections, the view- 
point that attitude measures necessarily predict behavior has been sub- 
jected to critical inquiry (16, 58a, 87), For example, according to Pace, 
"Many attitude tests are descriptive but not predictive, and their mean- 
ing and interpretation is limited by this fact. Definitions of attitude as a 
tendency to act may need to be reconsidered; acceptance of the definition 
implies that behavior is the criterion of validity" (82). 

As indicated in studies described in this section, and also in Chapter 4, 
the possibility of predicting behavior from attitude measures has not 
been discarded by -either the survey specialist or by the social scientist. 
It is recognized that the difficulty may be in the failure of attitude-sur- 
veying techniques to structure the situation in such a way as to tap those 
attitudes which are, in fact, indicative of the tendency to respond or 
perform in a particular way (90). Involved also is the distinct possi- 
bility of differences between the expressions of attitudes by the indi- 
vidual "in public" and those which appear in a more intimate and "pri- 
vate" situation. Such considerations have led to a centering of attention 
upon (a) the analysis of factors which can tend to make people's opin- 
ions and attitudes an adequate basis for predicting their behavior and 
(b) formulating steps which may be taken to improve the validity of 
opinion and attitudes as predictors of performance ( 32a, 33) . In addition, 
experimental studies, of the type described in earlier chapters, are help- 
ing to define the conditions under which, and the boundaries within 
which, behavior in industry can be predicted from attitude measures. 

Validity as a Matter of Interpretation: While prediction is of utmost 
importance, and is generally conceived as the major purpose of possibly 
most attitude surveys, the matter of interpretation may be construed as 
a criterion of validity in itself, particularly when prediction of behavior 
is not the prime purpose of a survey. In this sense, the problem of validity 
involves the query as to whether a question or series of questions elicit 
a response which reflects the individual's true opinion or attitude toward 
an issue in question, or is a reflection of his opinion or attitude on some 
other issue. 

As suggested by McNemar, this problem "boils down to whether a 
person's opinion or attitude can be safely inferred from what he says. Or, 
to state the problem differently, it can be said that validity is a matter 
of interpretation. Suppose a soldier says he would like to go overseas soon, 
and that this is interpreted as indicating a desire to get into active combat. 
At once the question of validity faces us. Does his statement really indi- 
cate high fighting morale, or does it merely indicate a desire for change 


or for adventure?" (72, p. 315) . In other words, the matter of interpreta- 
tion can influence prediction, in that an error in the interpretation of 
response will lead to error in predicting desired or anticipated perform- 
ance. Specifically, in the illustration cited by McNemar, an investigator 
may err in predicting combat performance from the question under con- 
sideration, while he might be able to predict some form of social behavior 
on shipboard. 

Whether or not opinion or attitude is related to behavior, there is still 
the fact that opinion or attitude can have "separate validity of its own" 
(83). The significance of this type of validity appears, for example, in 
studies concerned with the measurement of "job satisfaction." As Korn- 
hauser points out, "the proportion of the working population classified 
as dissatisfied will depend in a large measure upon the arbitrary method 
of defining what the term dissatisfaction refers to in the given case" (61, 
p. 132). This possibility is well illustrated in investigations conducted 
in Holland by the Netherlands Institute of Public Opinion. The per- 
centage of Dutch women and men replying "yes" to the question "Do 
you find satisfaction in the work you do?" was found to be approximately 
the same as in the case of Hoppock's investigation in the United States 
(50). However, further analyses of findings have led to the conclusion 
that "the concept of job satisfaction, as used to date, is a very confusing 
one. Our own feelings," states Stapel, "are . . . that we are measuring 
something in our respondents which is not necessarily related to the job 
they happen to have" (104). As appears in the section which follows, 
and also in Chapter 14, the matter of interpretation as a criterion of valid- 
ity is also a highly significant issue in the case of studies of employee 

In discussing the problem of validity of conventional opinion-attitude 
measures, in 1946, McNemar called attention to the pressing need for 
supplying evidence that "the device used measures or classifies the at- 
titude or opinion that it was designed to measure" and also evidence that 
"verbal or symbolic" responses can be depended upon to indicate an 
"individual's action tendency" (72, p. 296). Similarly, examination of 
studies in the specific area of employee-attitude surveying, described in 
later chapters, shows that, in spite of some notable exceptions, the prob- 
lem of validity continues to be a major one in the evaluation of outcomes 
with respect both to the matter of interpretation and the prediction of 


According to Katz, the most common weakness of attitude surveys, on 
the technical side, is a lack of adequate research design (57). Lack of 


adequate research design is particularly disadvantageous from the view- 
points of building up a field of knowledge and of arriving at generaliza- 
tions concerning the sources of motivation and morale. By survey design, 
according to Katz, "is meant the attempt to establish a relationship be- 
tween two factors" and most especially to reveal causal relationships 
among factors. Failure to do this is illustrated in the tendency, referred 
to earlier, of depending solely upon counting procedures where provisions 
should be made for the study of interrelationships. Neglect of careful 
planning of almost any feature of the survey can eliminate or at least 
limit the possibility of accomplishing this result. This can happen, for 
instance, when the survey is confined to the verbal responses of one sam- 
ple of respondents 'when multiple samplings are needed for a complete 
assessment of the situation. Thus an investigation which undertakes to 
survey supervisory attitudes should properly include provisions for obtain- 
ing information on the relation of such attitudes to differences among 
supervisors with respect to employee attitudes, on the one hand, and 
management behavior, on the other. A more limited approach can pro- 
vide answers as to how supervisors feel, but fails to provide clues which 
have practical value, even at the applied level, as to the source of such 
feelings. Furthermore, according to Katz, good survey practice calls for 
extensive and searching use of experiential questions and for the collection 
of outside (objective) criteria of satisfaction and morale. "A study will 
be definitely enriched if these background, factual items can be supple- 
mented with information about additional criteria . . . Significant prog- 
ress can be made if we give more attention to this factor of establishing 
relations between measures which are independently obtained" (57, p. 

The need for adequate research design is widely recognized, even in 
the case of surveys conducted at the applied level. Less attention has 
been given to the need for using survey designs in a theoretical context. In 
a critical review of opinion-attitude methodology, McNemar has pointed 
out that "the 'hypothesis' step in scientific reasoning and research seems to 
be all too frequently ignored" in opinion and attitude surveys (72). As 
indicated on page 223, the need for rationale, or theoretical context, is 
particularly crucial in surveys which are intended to yield broad principles 
and generalizations in the area of motivation and morale, rather than 
a series of facts which are of interest only in relation to the immediate 
needs of a particular company ( 57) . 

The need for theoretical context is illustrated by the variety of senses 
in which the term "morale" is used in employee-attitude surveys. 6 Treat- 
ment of this highly important concept varies from acceptance of the view- 

f^ r 1 4 


point that "morale" is what employees believe it to be to a posteriori 
definitions in terms of whatever the survey may show about the feelings, 
sentiments, and attitudes of employees. It is apparent that not alone 
the survey used as a research tool but also the practically oriented survey 
can bring firmer outcomes if it starts with an operational or conceptual 
definition of morale (71). Specific advantages are achieved if, as in the 
case of the definition given on page 284, this is derived from pertinent 
personality and social theory (102). 

As suggested earlier, the need for theoretical context becomes particu- 
larly apparent when attention is focused upon attitudes and behavior re- 
flecting the influence of the group. The survey method elicits individual 
responses from individuals acting alone (71); it makes the individual 
the unit of measurement (57). However, it has become increasingly ap- 
parent that behavior is frequently a group process and that motivation is 
frequently a phenomenon of social interactions within the group. This 
points to the need for adapting survey methods to the study of group 
attitudes. To do this involves consideration of concepts of group struc- 
ture to be derived from emerging psychological theory which yield 
fruitful hypotheses for research and point to the variables that need to 
be studied (57, p. 221). 


Perhaps the most treacherous shoals encountered in threading a course 
through employee-attitude-survey findings are in the interpretation of re- 
sults (57). Errors in interpretation arise, in part, from failure to consider 
and weigh technical deficiencies of the kind discussed in earlier sections of 
this chapter. In addition, there are the less tangible but at least equally 
important errors of inference which help account for the often wide dif- 
ferences in the interpretation of survey data. 

Overgeneralization: Such errors of inference arise frequently from the 
tendency to over generalize the findings of employee-attitude surveys. This 
appears particularly in the frequency with which investigators have gen- 
eralized the results obtained in a single survey, conducted in a particular 
situation, into broad conclusions concerning the significance of particu- 
lar factors in industry at large. Hazards also exist in arriving at generaliza- 
tions based upon a consideration of accumulated findings from independ- 
ent studies in so far as these differ in objectives, methodology, content, 
and other respects. Furthermore, the not infrequent appearance of con- 
tradictory findings in such studies places further limitations on the free- 
dom with which generalizations may be made. 

A not uncommon error in the way of Overgeneralization takes the form 
of conclusions concerning courses of action. For example, as Kornhauser 


(61) points out, the fact that 75 per cent of employees in a plant express 
dissatisfaction with lockers, and only 25 per cent are dissatisfied with 
wage rates, does not warrant the conclusion, at times formulated from 
such data, that management is justified in giving attention only to the 
former and ignoring the latter. Kornhauser furnishes another instance 
of overgeneralization by reference to a study (27) in which the employees 
were asked whether they had seen a statement of the company's financial 
situation and, at the same time, asked how much more money the com- 
pany should pay them per week (61). A comparison showed that the 
average wage increase desired by those who had seen the company's 
financial statement was $3.07; by those who had stated that they had 
not seen the statement, $5.53. These findings are quite unjustifiably 
interpreted as meaning that the "desired wage increase can be reduced 
almost by half by showing the company's financial statement to em- 

The danger of overgeneralization appears not only in conclusions bear- 
ing upon courses of action, but also in dealing with inferences concerning 
basic motives as derived from attitude surveys. For example, the fact 
that a large number of employees rate wages high in importance, or other- 
wise express attitudes indicative pf the desire for higher wages, has been 
alternatively interpreted as revealing a basic need for subsistence (food, 
clothing, etc.); a drive toward security; the desire for social status, or as 
reflecting the social motives of the work group. 

The psychologist is, unfortunately, faced with the necessity of making 
inferences concerning such motives because they cannot be observed 
directly. Any or all of the inferences referred to above may be justified, 
particularly when theoretical conceptions and related studies furnish 
the framework for interpretation and for the formulation of general con- 
clusions. However, a minimum requirement in the present state of 
limited knowledge concerning the basic mechanisms of human behavioi 
is at least to suggest alternative interpretations, rather than to limit 
generalizations to particular frames of reference or values which have 
the greatest appeal to the investigator. This is a problem which is per- 
haps of no great concern from the viewpoint of immediate application. 
It is, nevertheless, of tremendous consequence for increasing knowledge 
and building sound theory in the fields of industrial and social psychology, 
and in constructing a science of industrial relations. 

Oversimplification: Closely related to and at times overlapping the 
problem of overgeneralization is the tendency to oversimplify the situa- 
tion in interpreting survey findings. In fact, the practice of relating specific 
attitudes to specific motives, discussed above, might be viewed as over- 
simplification rather than overgeneralization. A more specific (and per- 


haps more acceptable) illustration of oversimplification is found in the 
process of verbal identification whereby a study becomes an investiga- 
tion of a particular kind simply through the literary equivalent of a 
"christening" ceremony. So, the 1947 National Industrial Conference 
Board survey, described in Chapter 13, which actually measured only 
the relative importance attached by employees to specific factors, be- 
comes identified as and therefore presumably is in fact a study of 
"factors affecting employee morale," although no external criterion of 
morale whatsoever was used in the investigation. 

Significance of the Industrial and Social Background: Of greater con- 
sequence and also more frequent is the tendency referred to in earlier 
pages of treating single variables as though they were actually operating 
with complete independence, and the related practice of treating ob- 
tained relationships as though they were causal in character. Such forms 
of oversimplified interpretation reflect, in part, deficiencies in experi- 
mental and statistical design. Of additional significance is the question 
of whether the influence of any incentive or determinant of morale, or 
the origin of any attitude and the nature of basic motives, can be fully 
assessed without reference to the broader industrial and social background 
within which employee attitude surveys are conducted. According to 
Kornhauser, "It seems clear that balanced judgments about the influ- 
ences that are important in each particular industrial situation, while 
they may well make use of whatever evidence is available from experi- 
ments and statistical analyses, need to go far beyond these. They must 
utilize a wide range of information about the specific company, the work- 
ing group, and the history of industrial relations in that concern . . . 
Further, however, the investigator can understand and interpret the local 
morale picture only if he is fully alive to the vastly significant influences 
which lie entirely outside the company social and economic conditions 
in the community and in the world at large" (61, p. 142) . 

Professional and Personal Biases: There can be no serious quarrel with 
the viewpoint that the psychologist concerned with the application of 
findings from both experiments and surveys cannot operate outside of 
the "social-historical setting." Quantitative findings from employee-atti- 
tude surveys (and also from experimental studies) have little practical 
value when considered in vacua. They acquire usefulness in dealing with 
the realistic problems of production, satisfaction, and morale only when 
projected against the background of broader economic and social de- 
velopments characterizing a situation or an era. This necessity, however, 
introduces the distinct possibility of errors in interpretation which are 
derived from the professional and personal bias of the investigator (31). 
The influence of such bias, or of what might be described as differences in 


the social perspectives of research personnel, is illustrated in the following 
citation from Buckingham 

'Two men let us call them curriculum experts are given the re- 
sults, in full statistical detail and reliable to the fast degree, of the survey 
of a school district in a mill town. Each of these men is asked to lay 
out the offerings, or at least to lay down principles, for the work of the 
schools in that district. Will these men agree in their findings? By no 
means. One of them may be a realist, who will say in substance, 'The 
information at my disposal is that 70% of these boys and girls are going 
into the mills and factories of this town. It is clear that they are going 
to need the things used in these occupations. Accordingly, let us teach 
them these very things. Let their English be industrial English. Let their 
mathematics be the mathematics of the forge and factory. Provide shops 
like those in which they will serve and cooperative part-time arrange- 
ments with the industries whose operatives they will become/ 

"The other curriculum expert may be a humanist. He may say 'These 
boys and girls are going to lead narrow lives unless we can give them 
inner resources. They will have enough of forge and factory. Let us give 
them in the schools before it is too late glimpses of wider horizons. Let 
them paint and draw and make music. Let them go adventuring into 
the past and into far countries. In English let us bring them in contact 
with beauty of form and nobility of thought and let us encourage them 
in self-expression. In short, in a fine humanistic sense, let us give them 

"These two curriculum experts are miles apart. Each was supplied 
with the same statistical facts, and neither of them questioned the truth 
of these facts. Moreover, neither of them brought to bear upon the prob- 
lem any additional statistical facts. Yet the curriculums which they 
recommend are quite different. Evidently, these men have interpreted 
the statistics each in his own way; and evidently too, the interpretations 
are powerful factors, for although resting upon the same grounds they 
have led to opposite conclusions . . ." (13, p. 382). 

It is evident that underlying this difference in interpretation is a differ- 
ence in "educational philosophy" a difference in social perspective con- 
cerning the place and the role of a public-school system in a highly in- 
dustrialized community. Another striking illustration of such effects upon 
interpretation of survey data is found in the reconsideration, by Ansbacher 
(3), of an investigation designed to examine the attitudes of foreign 
workers so-called "slave labor" in wartime Germany (112). A com- 
parison of responses given by French, Italian, and Russian workers led 
to the conclusion that the Russians were most anti-German. But, accord- 
ing to Ansbacher, the same data may be interpreted as showing that the 
Russians were actually more favorable to the Germans than were the 
other groups. This reinterpretation is arrived at, in part, by a re-examina- 
tion of the data which leads Ansbacher to conclude that some of the 
judgments made by the investigators were "forced and unconvincing/' 


More important still, the reinterpretation is based upon consultation 
with German psychologists and personnel men who had close contact 
with foreign workers during the war (4), and upon alterations in certain 
of the premises upon which interpretation is based. 

Ansbacher considers in some detail, and with commendable frank- 
ness, the question which constitutes a serious basic problem for social 
psychology of why the scientists of the Bombing Survey saw one mean- 
ing in data in which he later perceived an entirely different meaning. In 
doing so, he follows the theoretical formulations of Krech and Crutch- 
field (62), derived from research in the area of "social perception" dis- 
cussed in Chapter 5. 

From a reconstruction of the "cognitive structure of the Bombing Sur- 
vey psychologists" and of the "views which were widely held at the time 
they made their interpretation," Ansbacher reaches the conclusion that, 
in the reinterpretation of findings, ". . .we have a truly striking exam- 
ple that assimilation of new facts to existing beliefs 'is not a fault found 
among the prejudiced few. It is to be found universally and is due to 
the very nature of our perceptual and cognitive processes' (62). Even in 
the case of social psychologists, new perceptions will be 'distorted' and in- 
fluenced by the nature of the major cognitive structures" (3, p. 135). 

Reduced to simple essentials, this analysis of sources of professional 
and personal bias simply points to the fact that the outlook of even the 
social scientist as reflected in his interpretation of research data is in- 
fluenced by his education, experience, environment, and established 
point of view. The possible influence of "cognitive structures" has been 
noted in the discussion of the Hawthorne studies, of investigations con- 
ducted in the plant of the Harwood Manufacturing Co., and of other 
experiments discussed in earlier chapters. Similar references will be found 
in the discussion of the findings of attitude surveys in later chapters. 

In a review of a volume on Communications Research (65), Sandage 
calls attention to the fact that "the psychological and sociological at- 
tributes of the researcher at times color conclusions" (93, p. 239). For 
the professional psychologist, acquainted with the points of view and 
conceptual preferences of his colleagues, making allowances for pro- 
fessional biases in interpretation is not too difficult a matter. For the lay 
reader and others who are unfamiliar with the theoretical context favored 
by a particular investigator, the situation presents greater hazards. Even 
more troublesome is the problem of making allowances for the personal 
biases (as contrasted with bias in terms of theory and methods) which 
sometimes, in spite of the best intentions of the investigator, creep into 
the interpretation of findings. Perhaps each report in an area so com- 
plex as that of motivation and morale, conducted in an era of rapid 


economic and social change, and involving broad social issues, should 
carry a brief biography of the author and some statement of his social 
point of view. In this connection, it is probably very apparent that the 
personal preference of the present author is for a continuation of the 
"free enterprise" system, and that he looks primarily to enlightened man- 
agement for conscientious efforts and effective outcomes in applying the 
findings of research with the view of increasing production, individual 
satisfaction and morale in industry. Perhaps the author has failed to 
compensate sufficiently for his own personal bias to produce an alto- 
gether objective evaluation of studies in the field of motivation and 
morale. If such should appear to be the case it does not result, at least, 
from a failure to make a constant effort to do so. 

General Outcomes of Employee- Attitude 

The above discussion reveals a formidable list of considerations which 
need to be borne in mind in the evaluation of survey data. Fortunately 
it is seldom that any single study is characterized by all, or an extended 
multiplicity of, sources of variation and error to which reference has been 
made. Moreover, as Deming has pointed out, the errors in a survey are 
not always additive. "It should ^ot be inferred," he adds, "that there 
are grounds for discouragement or that the situation is entirely hopeless 
with regard to the attainment of useful accuracy" (31, p. 369) . In spite of 
one or another defect in individual studies, employee-attitude surveys 
have yielded a vast amount of useful information about employee atti- 
tudes and have provided a basis for provocative inferences concerning the 
psychological foundations of motivation and morale in industry (14). 

1. Employee-attitude surveys have thrown light on the wide variety 
of factors in the plant situation with which employees are concerned. 
They have provided at least a rough measure of the relative importance 
attached by employees to different aspects of the work situation and of 
work relationships. 

2. Employee-attitude surveys have made it possible to assess with 
some degree of accuracy the level of employee satisfaction and morale. 
They have helped to provide management with "a clear impression as 
to whether the mental health of the organization as a whole is reasonably 
good or shows disturbing weaknesses" (61, p. 130) . Through comparison 
of various plants, departments, occupational and other groups, manage- 
ment is placed in a position to take action in those units which require 
special attention. 

3. Employee-attitude surveys have contributed to the identification of 
the factors which have particular significance as incentives and as de- 
terminants of employee satisfaction and morale. 


4, The measurement of employee attitudes provides useful methods 
for checking inferences concerning the basic motives underlying per- 
formance and adjustment at work. In general, the survey method espe- 
cially as used in investigations designed in sound theoretical context and 
in connection with field experimentation (58) represents a useful tool 
in the extension of motivational theory and in the enhancement of the 
social sciences. 

In industry, as shown in Chapter 21, attitude surveys have brought 
important results in the way of improved working conditions; in strength- 
ening training programs for employees and supervisors; in putting super- 
vision "on its toes"; in promoting better understanding between man- 
agement and the unions, and in other ways (92). Incidental to such 
achievements, attitude surveys help relieve tensions by letting workers 
unburden themselves and improve morale by showing that management 
is really interested in the people on the job (61 ). 

In general, studies described in later chapters fully support the opin- 
ion, as stated by Kornhauser, that "findings almost always add substan- 
tially to management's previous information about its employees and 
very often contain startling surprises. Perhaps the greatest accomplish- 
ments have been in the way management is shocked out of its compla- 
cency with respect to employee satisfaction. Even where the results at 
first seem dubious to managers, further detailed inquiry into the situation 
usually reveals some genuine basis for the feelings expressed in the sur- 
vey a basis either in the objective conditions or in the subjective mean- 
ings these situations have been permitted to assume" (61, p. 130) . 

At the same time, perhaps the most important need in initiating and 
conducting a survey is an honest determination on the part of manage- 
ment to do something with the findings. Industrial relations may deteri- 
orate, rather than improve, if management conducts a survey and fails 
to do everything in its power to take positive action for eliminating the 
sources of dissatisfaction and low morale uncovered by a survey. As is 
pointed out by the National Industrial Conference Board, "this does 
not imply that a company must commit itself to grant every wish or 
make every change indicated in an attitude survey's results . . . Some 
employee requests and suggestions simply cannot be and should not 
be fulfilled. But in that case, the company does have the clear-cut re- 
sponsibility to explain fully and frankly the reasons for not taking action." 
Furthermore, according to the Conference Board, employee-attitude sur- 
veys cannot be used as "substitutes for good, direct, face-to-face contacts 
between management and the worker 7 . . . They supplement personal 
contacts. They can never take their place" (39, p. 14). 

T Not italicized in the original. 



1. R. L. Ackoff and L. Pritzker, "The methodology of survey research/' Int. /. Opin. 
Att. Res., 1951, 5, 313-34. 

2. G. W. Allport and L. Postman, The Psychology of Rumor. Henry Holt & Co., 

3. H. L. Ansbacher, 'The problem of interpreting attitude survey data/' Pub. 
Opin. Quart., 1950, 14, 126-38. 

4. H. L. Ansbacher, "Testing, management, and reactions of foreign workers in 
Germany during World War II," Amer. PsychoL, 1950, 5, 38-49. 

5. S. Barkin in L. R. Tripp (Ed.), Proceedings of the Fourth Annual Meeting, In- 
dustrial Relations Research Association. Madison, Wis.: (Dec.) 1951, 80-84. 

6. J. K. Bell, Protective Techniques. Longmans, Green & Co., Inc., 1948. 

7. R. E. Bernberg, "The direction of perception techniques of attitude measure- 
ments/' Int. J. Opin. Att. Res. f 1951, 5, 397-406. 

8. A. E. Blankenship, "Methods of measuring industrial morale" in G. W. Hart- 
mann and I. N. Newcomb (Eds.), Industrial Conflict, The Cordon Co., 1939, 

9. A. B. Blankenship, How to Conduct Consumer and Opinion Research. Harper 
and Brothers, 1946. 

10. A. B. Blankenship, "The influence of the question form upon the response in a 
public opinion poll/' Psychol. Rec., 1940, 3, 349-422. 

11. A. B. Blankenship, "The effect of the interviewer upon the response in a public 
opinion poll," /. Consult. Psychol., 1940, 4, 134-36. 

12. A. B. Blankenship, "A source of interviewer bias," Int. /. Opin. Att. Res., 1949, 
3, 95-98. 

12a. D. W. Bray, "The prediction of behavior from two attitude scales," /. Abn. 
Soc. Psychol., 1950, 45, 64-84. 

13. B. R. Buckingham, "Statistics and modern educational thought," School Soc., 
(March 22) 1930, 31, 379-85. 

14. R. Burns, "Employee morale its meaning and measurement" in L. R. Tripp 
(Ed.), Proceedings Fourth Annual Meeting, Industrial Relations Research Asso- 
ciation. Madison, Wis.: 1951, 52-68. 

15. D. Calahan and N. C. Meier, "The validity of mail ballot polls," Psychol. Res., 

16. A. Campbell and R. L. Kahn, The People Elect a President. University of 
Michigan Press, 1952. 

17. D. T. Campbell, "The indirect assessment of social attitudes," Psychol. Bull., 
1950,47, 15-38. 

18. H. Cantril (Ed.), Gauging Public Opinion. Princeton University Press, 1944. 

19. H. Cantril, The "Why" of Man's Experience. The Macmillan Co., 1950. 

20. R. B. Cattell, A Guide to Mental Testing. University of London Press, 1936. 

21. R. B. Cattell, "The ergic theory of attitude and sentiment measurement," Educ. 
and Psychol. Meas. t 1947, 7, 221-46. 

22. R. B. Cattell, A. B. Heist, P. A. Heist and R. G. Stewart, "The objective meas- 
urement of dynamic traits," Educ. Psychol. Meas. t 1950, 10, 224-48. 

23. R. B. Cattell, E. F. Maxwell, B. H. Light, and M. P. Unger, "The objective 
measurement of attitudes," Brit. J. PsychoL (Gen. Section), 1949, 40, 81-90. 


24. F. S. Chapin, 'Trends in sociomctrics and critique/' Sociometry, 1940, 3, 245-62. 

25. C. W. Churchman, R. L. Ackoff, and M. Wax (Eds.), Measurement of Con- 
sumer Interest. University of Pennsylvania Press, 1947. 

26. K. E. Clark and P. H. Kriedt, "An application of Guttman's new scaling tech- 
nique to an attitude questionnaire," Educ. Psychol. Meas., 1948, 8, 215-24. 

27. R. J. Cole, "A survey of employee attitudes," Pub. Opin. Quart., 1940, 4, 497- 

28. S. M. Corey, "Professed attitudes and actual behavior," /. Educ. Psychol., 1937, 
28, 271-80. 

29. L. P. Crespi, 4 The interview effect in polling," Pub. Opin. Quart., 1948, 12, 98- 

30. D. D. Day, "Methodological problems in attitude research/' /. Soc. Psychol., 
1941, 14 (First Half), 165-79. 

31. W. E. Deming, "On errors in surveys," Amer. Soc. Rev., 1944, 9, 359-69. 

32. W. E. Deming, Some Theory of Sampling. John Wiley and Sons, Inc., 1950. 
32a. S. C. Dodd, "Predictive principles in polls scientific method in public opinion 

research," Pub. Opin. Quart., 1951, 15, 23-42. 

33. J. Dollard, "Under what conditions do opinions predict behavior," Pub. Opin. 
Quart., 1948, 49, 12, 623-32. 

34. S. S. Dubin, "Verbal attitude scores predicted from responses to a projective 
technique," Sociometry, 1940, 3, 24-48. 

35. L. W. Doob, Propaganda, Its Psychology and Technique. Henry Holt & Co., 
1935. Public Opinion and Propaganda. Henry Holt & Co., 1948. 

36. A. L. Edwards and K. C. Kenney, "A technique for the construction of attitude 
scales," /. Appl. Psychol., 1948, 32, 374-84. 

37. E. L. Ehle, "Techniques for the study of leadership," Pub. Opin. Quart., 1949, 

38. C. E. Evans, "Item structure variation as a methodological problem in an em- 
ployee survey," Amer. Psychol., 1949, 4, 280. 

39. Experience with Employee Attitude Surveys (Studies in Personnel Policies No. 
115). New York: National Industrial Conference Board, 1951. 

40. H. J. Eysenck and S. Crown, "An experimental study in opinion-attitude meth- 
odology," Int. /. Opin. Att. Res., 1949, 3, 47-86. 

40a. J. J. Feldman, H. Hyman, and C. W. Hart, "A field study of interviewer effects 
on quality of survey data," Pub. Opin. Quart., 1951-52, 15, 734-61. 

41. L. K. Frank, Projective Methods. Springfield, 111.: Charles O. Thomas, 1948. 

41a. J. R. P. French, Jr., A. Kornhauser and A. Morrow (Eds.), "Conflict and co- 
operation in Industry," /. Soc. Issues, (February) 1946, 2. 

42. J. G. Franz, "Survey of sociometric techniques with an annotated bibliography," 
Sociometry, 1939, 2, 76-92. 

43. J. P. Guilford, Psychometric Methods. McGraw-Hill Book Co., 1936. 

44. L. Guttman, "The Cornell technique for scale and intensity analysis" in C. W. 
Churchman et al. (Eds.), The Measurement of Consumer Interest. University 
of Pennsylvania Press, 1947, 60-85. 

45. L. Guttman, "The quantification of a class of attributes: A theory of a method 
of scale construction" in P. Horst et al., The Prediction of Personal Adjustment. 
New York: Social Science Research Council, 1941, 319-48. 

46. V. Hammond, "Measuring attitudes by error choice," /. Abn. Soc. Psycho/., 
1948, 48, 38-49. 


47. M. Haire and J. S. Gottsdanker, "Worker adjustment as studied by projectivc 
and interviewing techniques," Amer. Psychol, 1949, 4, 262. 

48. R. B. Hersey, "Emotional cycles in man," /. Ment. Set., 1931, 77, 151-69. 

49. R. B. Hersey, Workers' Emotions in the Shop and Home. University of Pennsyl- 
vania Press, 1932. 

50. R. Hoppock, Job Satisfaction. Harper and Brothers, 1935. 

51. E. L. Horovitz and R, Horovitz, "Development of social attitudes in children/' 
Sociometry, 1938, 1, 301-38. 

52. M. Jahoda, M. Deutsch, and S. W. Cook (Eds.), Research Methods in Social 
Relations (2 volumes). The Dryden Press, 1951. 

53. J. G. Jenkins, "Nominating technique as a method of evaluating Air Group 
morale," /. Aviat. Med., 1948, 19, 21-29. 

53a. J. G. Jenkins, Staff of Aviation Psychology Branch USN., E. S. Ewart and J. B. 
Carroll, The Combat Criteria in Naval Aviation (NRG Committee on Aviation 
Report No. 6) . Washington: US Navy, BuM and S, Division of Aviation Medi- 
cine, 1950. 

54. W. B. Johnson, "Euphoria and depressed moods in normal subjects," Character 
and Pers., 1937, 6, 79-98, 188-202. 

55. A. W. Jones, Life, Liberty and Property. J. B. Lippincott Co., 1941. 

56. D. Katz, "Do interviewers bias poll results?" Pub. Opin. Quart., 1942, 6, 248-68. 

57. K. Katz, "Good and bad practices in attitude surveys in industrial relations" in 
Proceedings, Second Annual Meeting IRRA. Champaign, 111.: Industrial Rela- 
tions Research Association, 1950, 212-21. 

58. D. Katz, "Survey techniques in evaluation of morale" in J. G. Miller (Ed.), 
Experiments in Social Process. McGraw-Hill Book Co., 1950, 65-77. 

58a. D. Katz, "The public opinion polls and the 1940 election," Pub. Opin. Quart., 
1941,5,52-78. * * 

59. A. Kornhauser, "Are public opinion polls fair to organized labor," Pub. Opin. 
Quart., 1946, 10, 484-500. 

60. A. Kornhauser, "Industrial psychology as management techniques and as social 
science," Amer. Psychol, 1947, 2, 224-29. 

61. A. Kornhauser, "Psychological studies of employee attitudes," J. Consult. Psy- 
choi., 1944, 8, 127-43. 

62. D. Krech and R. S. Crutchfield, Theory and Problems of Social Psychology. 
McGraw-Hill Book Co., 1948. 

63. R. T. La Piere, "Attitudes vs. action," Soc. For., 1934, 13, 230-37. 

64. C. E. Evans and La V. N. Laseau, My Job Contest (Personnel Psychology Mono- 
graph No. 1). Washington: Personnel Psychology Inc., 1950. 

65. P. F. Lazarsfeld and F. Stanton, Communications Research, 1948, 1949. Harper 
and Brothers, 1950. 

66. D. Lerner, "Public receptions of 'The American Soldier' " in R. K. Merton and 
P. F. Lazarsfeld (Eds.), Continuities in Social Research: Studies in the Scope 
and Method of "The American Soldier" Glencoe, III: Free Press, 1950, Chap- 
ter 7. 

67. J. M. Levine and G. Murphy, "The learning and forgetting of controversial 
material," /. Abn. Soc. Psychol., 1943, 38, 507-17. 

68. R. A. Likert, "A technique for the measurement of attitudes." Arch. Psycho/., 
1932,22,1-55. ' 

69. H. C. Link and A. D. Freiberg, "Is Dr. Kornhauser fair to organized pollsters?" 
Pub. Opin. Quart., 1947, 11, 198-204. 


70. H. C. Link and A. D. Freiberg, 'The problem of validity vs. reliability in public 
opinion polls," Pub. Opin. Quart., 1942, 6, 87-98. 

70a. C. P. Loomis and H. G. Pepinsky, "Sociometry, 1937-1947: Theory and meth- 
ods/' Sociometry, 1948, 11, 262-86. 

71. J. W. McConnell in Proceedings of Second Annual Meeting, Industrial Relations 
Research Association. Champaign, 111.: Industrial Relations Research Assn., 
1950, 222-25. 

72. Q. McNemar, "Opinion-attitude methodology," Psychol Bull, 1946, 43, 289- 


73. D. Manheimer and H. Hyman, "Interviewer performance in area sampling," 
Pub. Opin. Quart., 1949, 13, 83-92. 

74. J. L. Moreno, Foundations of Sociometry (Sociometry Monograph No. 4). 
Beacon House, 1941. 

75. J. L. Moreno, "Experimental sociometry and the experimental method in 
science," in Current Trends in Social Psychology. University of Pittsburgh Press, 
1948, 119-62. 

76. J. L. Moreno, Who Shall Survive? (Mental and Nervous Disease Monograph 
Series No. 58) . Washington: Mental and Nervous Disease Publishing Co., 1934. 

77. H. A. Murray and C. D. Morgan, "A Clinical study of sentiments," Gen. Psychol. 
Monog., 1945, 1. 1-154; II. 155-309. 

78. H. A. Murray, Explorations in Personality. Oxford University Press, 1938. 

79. T. E. Neely, A study of error in the interview. (Privately published) 1937. 

80. T. Newcomb, "Labor unions as seen by their members: An attempt to measure 
attitudes" in G. W. Hartmann and T. Newcomb (Eds.), Industrial Conflict. 
The Cordon Co., 1939, 313-38. 

81. R. C. Nickols, "Verbi," Int. J. Opin. Att. Res., 1949-50, 4, 574-86. 

82. C. R. Pace, "Opinion and action: A study in validity of attitude measurement/' 
Amer. Psychol., 1949, 4, 242. 

83. H. J. Parry and H. M. Crossley, "Validity of responses to survey questions/' 
Pub. Opin. Quart., 1950, 14, 61-80. 

84. M. B. Parten, Surveys, Polls and Samples. Harper and Brothers, 1950. 

85. S. L. Payne, "Thoughts about meaningless questions/' Pub. Opin. Quart., 1950- 
51, 14, 687-96. 

86. H. M. Proshansky, "A projective method for the study of attitudes/' /. Aim. Soc. 
Psychol, 1943, 38, 393-95. 

87. "Report on the analyses of pre-election polls and forecasts," Pub. Opin. Quart., 
1948-49, 12, 599-662. 

88. C. F. Reuse, "Differences between persons responding and not responding to 
a mail questionnaire," Amer. Sociol. Rev., 1943, 8, 433-38. 

89. S. A. Rice, "Contagious bias in the interview," Amer. J. Soc., 1929, 35, 420-423. 

90. D. Riesman and N. Galzer, "The Meaning of opinion/' Pub. Opin. Quart., 
1948-49, 12, 633-648. 

91. A. M. Rose, Union Solidarity. University of Minnesota Press, 1952. 

92. "Sampling workers' opinion," (J. W. Irwin) Dun's Review, (November) 1945, 
32 ff. 

93. C. H. Sandage, Int. /. Opin. Att. Res., 1950, 4, 239. 

94. F. H. Sanford, "The use of a projective device in attitude testing/' Pub. Opin. 
Quart., 1950-51, 14, 697-709. 

95. F. H. Sanford, The Use of a Projective Device in Attitude Surveying (Report 


No. 5). Haverford College and the Institute for Research in Human Relations, 

96. F. H. Sanford and I. M. Rosenstock, Protective Techniques on the Doorstep 
(Report No. 7). Haverford College and the Institute for Research in Human 
Relations, 1950. 

97. P. B. Sheatsley, "An analysis of interviewer characteristics and their relation- 
ship to performance/' Int. /. Opin. Att. Res., 1950-51, 4, 473-98; 1951, 5, 79- 
94, 191-220. 

98. M. Sherif and H. Cantril, The Psychology of Ego-Involvements. John Wiley 
& Sons, Inc., 1947. 

99. M. A. Sherif and H. Cantril, "The psychology of 'attitudes/ " Psycho/. Rev., 
1945, 52, 295-319; 1946, 53, 1-24. 

100. M. Sherif, "A study of some social factors in perception/' Arch. Psycho!., No. 

101. F. K. Shuttleworth; "Sampling errors involved in incomplete returns to mail 
questionnaires/' /. Appl Psycho/., 1941, 25, 588-91. 

102. R. G, Smith and R. J. Westen, Studies of Morale Methodology and Criteria (Re- 
search Bulletin 51-29). San Antonio: USAF Air Training Command, Human 
Resources Research Center, 1951. 

103. F. Stanton, "Notes on the validity of mail questionnaire forms/' /. Appl. Psycho/., 

104. J. Stapel, "What is job satisfaction?" Pub. Opin. Quart., 1950, 14, 551-54. 

105. H. Stember and H. Hyman, "How interviewer effects operate through question 
form," Int. /. Opin. Att. Res., 1949-50, 3, 493-512. 

106. J. S. Stock and J. R. Hochstim, "A method of measuring interviewer variability/' 
Pub. Opin. Quart., 1951, 15, 327-34. 

107. S. A. Stouffer, L. Guttman, E. A. Suchman, P. F. Lazarsfeld, S. A. Starr and 
J. A, Clausen, Measurement and Prediction (Vol. IV Studies in Social Psychology 
in World War II). Princeton University Press, 1950. 

108. E. A. Suchman and B. McCandless, "Who answers questionnaires," J. Appl. 
Psycho/., 1940, 24, 758-69. 

109. L. L. Thurstone and E. J. Chave, The Measurement of Attitudes. University of 
Chicago Press, 1929. 

110. A. B. Udow, "The 'interview-effect' in public opinion and market research sur- 
veys/' Arch. Psycho/., No. 277, 1942, 1-36. 

111. R. S. Uhrbrock, "Attitudes of 4430 employees," /. Soc. Psycho/., 1934, 5, 365-77. 

112. U. S. Strategic Bombing Survey, Morale Division, The Effects of Strategic Bomb- 
ing on German Morale Vol. II. U.S. Government Printing Office, 1946. 

113. M. S. Viteles, "The measurement of employee attitudes" in C. W. Churchman 
et d. (Eds.), Measurement of Consumer Interest University of Pennsylvania 
Press, 1947, 177-97. 

114. W. A. Wang, "Suggested criteria for writing attitude statements," I. Soc. Psy- 
cho/., 1932, 1, 367-72. 7 

115. I. R. Weschler, "An investigation of attitudes toward labor and management 
by means of the error-choice method I.," /. Soc. Psycho/., 1950, 32 (First Half), 
51-62. "A follow-up study on the measurement of attitudes toward labor and 
management by means of the error-choice method II.," /. Soc. Psycho/., 1950, 
32 (First Half), 63-67. 

116. I. R. Weschler, "Problems of indirect attitude measurement," Pub. Opin. Quart., 


117. I. R. Weschler and R. E. Bernberg, "Indirect methods of attitude measure- 
ment," Int. ]. Opin. Att. Res., 1950, 4, 209-28. 

118. "What the factory worker really thinks about productivity, nationalization of 
industry, and labor in politics," Factory Mgmt. Mainten., (January) 1946, 104, 
1, 80-88. 

119. C. Wedell and K. U. Smith, "Consistency of interview methods in appraisal of 
attitudes," /. Appl. Psychol., 1951, 35, 392-96. 

120. R. S. Woodworth, "Psychological experience with the interview," /. Pers. Res., 
1925,4, 164-165. 

121. J. C. Worthy, "Discovering and evaluating employee attitudes," Personnel Series 
No. 113, American Management Assn., 1947, 13-21. 


The Employee Looks Over 
the Plant Situation 

ONE MAJOR outcome of employee-attitude surveys has been 
to reveal the large variety of features of the plant situation with which 
employees are concerned. Surveys have also provided at least a rough 
measure of the relative importance attached by employees to different 
aspects of the work situation and x of work relationships (6) . 

The Hawthorne Plant Survey 

The fact that employee attitudes relate to many varied features in the 
plant situation, and the differences in values attached to individual items 
such as pay, working conditions, etc., can be seen by reference to three 
large-scale investigations. Among these is a survey of employee attitudes 
conducted in the Hawthorne Plant of the Western Electric Company. 
Observations made in the course of the original Relay Assembly Room 
Study, described in Chapter 10, suggested that one possible way of im- 
proving workers' attitudes and morale throughout the plant was by ask- 
ing employees to express their opinions on the conditions in their work- 
ing environment which they might like or dislike ( 12, 13, 19) . 

Such a study was undertaken in 1928, in connection with the develop- 
ment of a training course for supervisors which was being conducted in 
the plant at that time. The approach was first tried experimentally in 
the Inspection Organization, including about 1,600 skilled and unskilled 
employees in shops and offices. Upon the completion of the preliminary 
investigation in 1929, it was extended to other units of the plant to the 
point where by 1931, when the program was temporarily discontinued 
because of the economic depression, over 21,000 employees, representing 
about half the work force, had been personally interviewed (11) . 


In conducting the interviews used in the survey, each worker was ap- 
proached individually and asked to express his views. The employee was 
assured that he was invited and not ordered to express himself, for it 
was felt that voluntary comments would be the most reliable. It was 
also made clear that the content of the interview would remain entirely 
confidential, that neither names nor reference numbers were to be at- 
tached to the interviews, and that identifying statements which might 
reveal the employee or his location were not to be recorded. 

During the first part of the study, interviewers started with a set of 
questions which they expected to have answered by every one. These 
were not in the nature of a formal questionnaire, but the questions pat- 
terned the general field of approach to what the interviewer thought 
important. Following several modifications in interviewing techniques, 
the plan was finally adopted of permitting the employee, after an ex- 
planation of the program, to choose his own topic for discussion. This 
method, identified as the conversational or indirect approach, and cur- 
rently known as the nondirective approach (14, 15), has the advantage, 
according to the investigators, of stimulating the feeling of confidence, 
of giving the employee the benefits of emotional release, and of /being 
attention upon that which the employee considers important. 

An analysis and classification was prepared of comments voluntarily 
made by workers. At first only unfavorable comments were sorted out, but 
later both favorable and unfavorable comments were compiled for analy- 
sis. A classification of the content of interviews conducted in the Operat- 
ing Branch, in 1929, presented in Figure 20, illustrates the variety of 
features of the plant situation to which employees voluntarily referred 
in the nondirective interviews, and the frequency of mention of each 

Data of the type given in Figure 20 are ordinarily interpreted as provid- 
ing an indication of the degree of favorableness or unfavorableness of atti- 
tude toward specific objects, policies and other plant features. Applying 
this approach, it would seem necessary to draw the conclusion that em- 
ployees in the Operating Branch of the Western Electric Company were 
(in 1929) much concerned with wages and were generally dissatisfied with 
payments received for their work. This method of analysis was, indeed, 
applied to the comments on supervision which, as shown in Figure 20, 
included 2,770 unfavorable references in a total of 4,662 voluntary com- 
ments concerning supervision. In fact, in a preliminary report on the 
Hawthorne study, the conclusion was reached that "the relationship 
between first-line supervisors and the individual workers is of more im- 
portance in determining the attitude, morale, general happiness, and 
efficiency of the employee than any other factor" (12, p. 325). However, 










S642 00 ' 5 2468 

|80m) 60 

















60 120 


Figure 20. Classification of the Content of Interviews * 

this type of analysis was discarded in favor of the more "subtle" cer- 
tainly more subjective form of analysis and interpretation which has 
generally characterized interpretation of the Hawthorne data by the 
Harvard group. Thus, in referring to Figure 20, Mayo says 

* By permission, from E. Mayo, The Human Problems of an Industrial Civilization. 
Harvard Business School, Division of Research, 1946. (First printed by The Mac- 
millanCo., 1933). 


"Interesting as this chart is, it cannot, however, be regarded as more 
than a gallant attempt to analyze. According to those who were engaged 
upon the work, they could cull fragments only from each continuous 
interview; the continuity itself, the personal and more interesting part, 
eluded their grasp and escaped. I included the diagram here not merely 
for its own interest which is considerable, but also because it gives point 
to my claim that the interviewing group was beginning to realize that 
the interview experiment, like the test room, had stumbled upon a group 
of facts most important for industry, but facts exceedingly difficult to 
define. Interviewers were uneasily aware that in some indefinable way 
all was not well with the human situation in the Works. 

'The situation at first disclosed itself somewhat after this fashion. It 
was found that comments on material conditions of work had a higher 
validity than comments on persons. In certain instances a number of 
persons working in the same vicinity complained of smoke and fumes, or 
of cold, or of insufficient locker space, or of some other source of irrita- 
tion. Many of these complaints were found to be just and the specified 
condition was corrected. But complaints about persons, or for that mat- 
ter about supervision, in the great majority of instances had to be dis- 
regarded. Such complaints served indeed to draw attention to some- 
thing of interest in the personal history and present attitude of the 
person being interviewed; but the validity of the external reference was 
minimal" (11, p. 95). 

Such considerations led Mayo to discard the earlier conclusion con- 
cerning the prime importance of supervision and to conclude instead: 
"The interviewing programs showed that the major difficulty was no mere 
simple error of supervision, no easily alterable set of working conditions; 
it was something more intimately human, more remote. 'Supervision/ 
indeed, had shown itself to be another word which meant so many things 
that it meant nothing. In every department there was a human situation, 
these situations were never identical and in every different situation 
the supervisor played a different part" (11, p. 97). The general con- 
clusion reached by the Hawthorne investigators from an analysis of the 
survey findings is that interview data specifically complaints voiced dur- 
ing interviews cannot properly be "treated as facts in themselves but as 
symptoms or indications of personal or social situations which need to be 
explored" (13, p. 267). Thus, for example, 

"Suppose a worker, B, complains that the piece rates on his job are 
too low. In the interview it is revealed that his wife is in the hospital 
and that he is worried about the doctor's bill he has incurred. In this 
case, the latent content of the complaint consists of the fact that B's 
earnings, due to his wife's illness, are insufficient to meet his current 
financial obligations. This source of his dissatisfaction can be expressed 
in many different ways, one of which is to grumble about the piece 
rates" (13, p. 267). 


According to Roethlisberger and Dickson, the above case illustrates 
the necessity of differentiating between the manifest or the material 
content of the complaint, as contrasted with the latent content or the 
psychological form of the complaint. Emphasis upon the latent meanings 
led directly to the program of employee counseling initiated in the 
Western Electric Company in 1936 (13, 3, 20) and to the adoption of 
similar programs by other industrial organizations and in government 
service (1, 7, 16, 2). There is no question that interviews, used for this 
purpose, have therapeutic value "in relieving emotional strains and con- 
flicts" (10). However, the usefulness of interviews for this purpose pro- 
vides no basis for the neglect of data such as those cited in Figure 20 as 
indicators of attitudes toward specific factors in the plant situation. 

National Industrial Conference Board 

1947 Survey of Six Plants 

and "My Job Contest" 

Studies conducted more recently by the National Industrial Con- 
ference Board (5) and by the General Motors Corporation (4) furnish 
further striking evidence that employee attitudes and sentiments relate 
to a large variety of features of the work situation. In a survey conducted 
in 1947 by the National Industrial Conference Board, each employee in 
six manufacturing plants was asked to go through the list of 71 morale 
factors reproduced in Table 17 and to select and check the one that he 
felt was most important to him. He was then instructed to go through 
the list again and select the second most important factor. This process 
was repeated until the employee had made five selections. The form was 
then sent unsigned to the National Industrial Conference Board in a self- 
addressed stamped envelope supplied by the Board. The frequency with 
which employees in the six co-operating companies * rated each factor as 
first in importance is shown in Table 17. In Table 18 is shown the fre- 
quency with which each factor was placed by employees among the top 
five factors. 


The survey conducted by the National Industrial Conference Board 
was highly directive in character, in as much as the employees were 
limited to the rating of factors in a list prepared in advance by the in- 
vestigators. The procedure employed by the General Motors Corpora- 

* The number of nonsupervisory employees ranged from 190 in the smallest plant to 
2,300 in the largest. The percentage of returned questionnaires varied from 11.7 to 
53.2. Employees in five of the plants were represented by collective bargaining agree- 


tion, involving letters written by employees in a competition for prizes 
known as My Job Contest, was designed to be less directive in character, 
providing an opportunity for employees to project their attitudes, senti- 
ments and feelings with respect to their jobs and surrounding conditions. 
According to those responsible for the General Motors Corporation 
study, "survey techniques using specific questions limit sharply the 
factors which the respondent is to consider/' They direct attention to 
specific aspects of the work situation, whereas the worker really may be 
more concerned with others. They tend also, it is claimed, to center at- 
tention on unsatisfactory job conditions and on workers' "gripes." They 
"invite an airing of these gripes rather than provide evidence of the work- 
er's attitude toward the complete work situation" (4, p. 2). 

Such shortcomings in the conventional questionnaire techniques for 
surveying employee attitudes, combined with the desire to integrate such 
a survey with a long-term communication program, led to the decision, in 
1947, to use a letter-writing contest as a means for discovering what's on 
the worker's mind in the widely scattered plants of the General Motors 
Corporation. The subject finally selected for the contest, from which 
supervisory and management employees were excluded, was My Job and 
Why I Like It. Incentives towards participating and towards effective 
effort in writing letters were provided in the form of 5,145 prizes, valued 
at $150,000, including a Cadillac, Series 62, Convertible Coupe as a first 
prize, 49 other automobiles, 25 electric ranges, 50 automatic washers, 100 
portable radios, 750 pairs of guide spot lights, etc. 

An important provision of the contest was that employees would be 
free to discuss any or all favorable aspects of their jobs. Although the 
statement of the topic (My Job and Why I Like It) limited responses 
to the positive aspects of the job, it was felt by the investigators that 
important negative aspects would become conspicuous through the lack 
of mention given them. Moreover, an opportunity was furnished for em- 
ployees to include constructive criticism on a Post Script which would 
not be considered in contest judging. Underlying this approach was the 
desire to focus the employee's attention upon the positive aspects of 
his job, this being, in fact, the first of the following four major objectives 
of the contest, which was envisaged as a step in a total communication 
project, rather than only as a means of finding out what employees were 
thinking in specific areas of work satisfaction: 

"1. To encourage more constructive attitudes in the mind of em- 
ployees by directing their attention to the positive aspects of their jobs. 

"2. To place certain educational bulletins in the hands of employees 
that would indicate some of the benefits derived from employment with 
General Motors. 



TABLE 17. Morale Factors: National Industrial Conference Board Survey 

Percentage of Employees Rating Each Factor as First in Importance 

(after Factors Affecting Employee Morale, Studies in Personnel Policy 

No. 85, 1947) 

[Absence of per cent figure signifies less than ,1 per cent] 




Per cent selecting 

factor as 
first in importance 



Per cent selecting 

factor as 
first in importance 




1. Vacation and holiday practices 3.3 

2. Leave of absence practices 
(non-military) .5 

3. Job security employment sta- 
bilization 30.6 

4. Employee merit or perform- 
ance rating (an organized and 
systematic method of apprais- 
ing your performance) 2.5 

5. Practice of informing you of 
your job status (both of your 
success and failures) 3.1 

6. Personnel counselling .8 

7. Compensation (base pay) 8.7 

8. Formal plans for determining 
basic rates of pay (job evalua- r % 
tion programs) 1.0 

9. Extra compensation plans (all 
types of bonuses) .5 

10. Profit-sharing plans (excluding 
employee savings plans) 3.8 

11. Employee savings and thrift 
plans 1 .0 

12. Training of wage earners 
(rank-and-file employees) .4 

13. Training of supervisors and 
foremen .8 

14. Training of executives .2 

15. Induction training (including 
orientation) .1 40 

16. Company medical and health 
programs 1.3 

16a. Accident prevention activities .5 

17. Employee feeding facilities 
(restaurants, lunch rooms, caf- 
eterias, canteens, mobile kitch- 
ens) 1.1 

18. Total hours worked per day 

per week 1.1 

19. Shifts (time starting and fin- 
ishing work) .9 

20. Transportation facilities (to 

and from work) .5 

21. Housing facilities 

22. Smoking privileges 

23. Physical working conditions 

24. Physical condition of recrea- 
tion rooms, lavatories and 
other personal facilities 

25. General outside physical ap- 
pearance of plant 

26. Rest periods 

27. Length of lunch period 
Frequency of paying wages 
Employee financial benefits, 
such as group life insurance, 
sickness insurance, and pen- 

31. Social and recreational activi- 

Termination allowances (dis- 
missal compensation) 
Labor unions 
Type of union leadership 

36. Employment records accu- 
racy and completeness of em- 
ployee's history with the com- 

37. Music on the job 

38. Employment of mixed nation- 

39. Employment of mixed races 
Labor turnover rate (its effect 
on morale of employees) 
Opportunities in the company 
for advancement 

Practice of hiring outsiders for 
responsible jobs vs. "promotion 
from within" 

43. Type of company's product 
(its social importance) DUR- 

43a. Type of company's product 
(its social importance) DUR- 

44. Knowledge of company's prod- 

















Per cent selecting 

factor as Factor 

first in importance No. 

45. Knowledge of competitors' 

46. Knowledge of company's cus- 

47. Company's reputation with 
the public 

48. Company's reputation with its 

49. Your knowledge of company's 

50. Type and condition of tools 
and equipment 

51. Employee suggestion systems 

52. Methods of handling griev- 

53. Contact with executives (op- 
portunity to see them occa- 

54. Your burden of state and fed- 
eral taxes 

55. Marital status (single or mar- 

56. Domestic relations and home 

57. Employment methods, includ- 
ing selection, interviewing and 

58. Laws affecting relations be- 
tween employers and employ- 

59. Policies and practices regard- 
ing discharge of employees 




Per cent selecting 

factor as 
first in importance 

60. Policy with respect to wearing 
uniforms (or regulating work- 
ing attire) 

61. Production drives: 

(a) Based upon cooperative 
1 .0 effort of management and 

employees (WARTIME) .3 

2 (b) Incentive ideas stimulated 

by company (PEACETIME) .1 

62. Location of plant (on "right" 
or "wrong" side of "the 

.2 tracks") .1 

63. Your own temperament 
ability to get along with others 1.5 

.1 64. Your confidence in yourself 1.0 
65. Your family's attitude toward 
your company and job 

.5 66. Company's attitude toward 
employees (its interpretation 
of policies whether liberal or 
conservative) 2.3 

68. Supervisors' temperament and 
personality 3.5 

69. Bulletin boards, house organs 
and other methods of dissem- 
inating information to employ- 
ees .1 

70. Quality of supervision 1.3 

71. Type of work 7.2 

44 3. To collect material for the enlightenment and education of super- 
visory and management groups. 

"4. To obtain a body of data for the analysis of employee-attitudes/' 
(4, p. 4) 

Official entry blanks were obtained by employees from their immedi- 
ate foremen or supervisors who played an important role in stimulating 
employee participation. They also gave contestants all the help they 
could, including copies of the rules, a broadside picturing the prizes, and 
three "thought-starter booklets." One of these, Getting Started, provided 
helpful suggestions for "breaking-the-ice" and getting over the hurdle of 
writing an entry. Another outlined the Six Point Objectives of General 
Motors as stated by C. E. Wilson, President, in pointing out the job 
ahead for General Motors in the post-war period. The third, entitled 


U.S., The Road to Better Living, presented a broad statement about the 
advantages of life under the American System. 

Entry blanks were all numbered in series. Each contestant signed his 
name only on a detachable coupon identifying himself by General Motors 
plant, unit, department, badge number, home address, city. No identifica- 
tion other than serial number appeared on the contest letter itself, which 
was mailed to a Chicago Post Office box belonging to the firm retained to 
do the preliminary judging. 

Of the 297,401 eligible employees in some 49 cities throughout the 
United States, a total of 174,854 entered letters during the six-week con- 
test period, which closed on October 31, 1947. Analysis of entries showed 
that they represented 55.4 per cent of hourly and 823 per cent of salaried 
employees, totaling 58.8 per cent of eligible General Motors employees. 
Supplementary studies conducted by the Opinion Research Corpora- 
tion, involving three divisions, led to the conclusion that "there were 
no essential differences between the employees who participated in M/C 
[My Job Contest] and those who had not" (4, p. 32). The survey also 
indicated that (1) the contest was even more popular with employees 
than the unusually high returns might indicate; (2) upwards of three- 
quarters of the employees in the three plants studied ascribed good mo- 
tives to the company in holding the contest, only one-tenth subscribing 
to the "erroneous idea" that it was conducted "just to advertise General 
Motors Products"; though job morale had considerable effect on the 
individual's decision to enter the contest, the high percentage of entries 
could be largely attributed to promotion of the contest. 

All letters were read and evaluated by a staff of specifically trained 
readers under the direction of Lloyd Herold, of the Lloyd Herold Co., 
with the advice of several leading university specialists and a Board 
of Honorary Judges. On the basis of three M/C standards sincerity, 
originality, and subject matter 5,145 prize winners were selected. In 
addition, the two hundred letters rated as best were ranked by a Board 
of Honorary Judges, including a noted author and economist (Peter 
Drucker); a poet and newspaperman (Edgar A. Guest); a businessman 
and dean of a university school of commerce (James A. McCarthy); an 
educator noted foi his achievement in labor relations (George W. Tay- 
lor), and the U.S. Commissioner of Education (John W, Studebaker). 


With the completion of the contest and the award of the prizes, atten- 
tion was turned to the analysis of the data. Such analysis was undertaken 
by Evans and Laseau with the purpose of achieving the third and fourth 
objectives of M/C, viz., (1) to collect material for the enlightenment 


and education of supervisory and management groups and (2) to ob- 
tain a body of data for the analysis of employee attitudes. Underlying 
the analysis was the premise that "a human and personal document, such 
as employees submitted in MJC, is a record of a person's thoughts, when 
his mind is at liberty to discuss subject matter of interest or importance 
to himself '(4, p. 33). 

According to the investigators, while it is reasonable to assume that 
the typical contestant gave attention to aspects of his experiences which 
might influence the judges to consider his entry favorably, he also re- 
tained a high degree of mental freedom regarding what he would write 
about. 'This/' it is said, "is one of the most important aspects of MJC 
as a technique, for it produces a state of mind that is open and undi- 
rected. The technique elicits for consideration all sorts of ideas, experi- 
ences, and theories on the part of the individual. In this unconfined state, 
concepts of importance or interest tend to float to the surface . . ." (4, 
p. 16). 

According to Evans and Laseau, like the techniques of psychoanalysis 
and nondirective counseling, MJC established a climate in which the indi- 
vidual was free to express his unconscious desires and reveal his inner 
personality. As a result, it could be assumed that an effective analysis 
of the MJC entries would provide highly reliable indications of the most 
important thoughts of General Motors' employees regarding their jobs 
and related experiences resulting from their association with General 

Methods of Content Analysis: Information concerning "employee 
thinking" was obtained by means of a content analysis of the letters in 
terms of themes discussed by the entrants. A preliminary list of themes 
was derived from 1,000 letters, representing every tenth entry from the 
first 10,000 letters. Copies of the 1,000 sample entries were read by five 
independent groups. Each submitted a list of themes to be coded as a 
basis for letter-content analysis. This yielded a total of over 150 topics 
or coding categories. Examination of the list, with particular reference 
to frequency of recurrence in the sample, made it possible to reduce the 
number of coding categories (thematic codes) by approximately one half. 
Further analysis suggested that not all of these were sufficiently sensi- 
tive to be useful in reporting results, and the number of themes employed 
in analyzing and reporting findings was reduced to 58. Both the char- 
acter of the letters written by employees and the nature of the thematic 
coding structure are illustrated in the sample entry with annotated theme 
mentions reproduced on page 265. 

Responsibility for analyzing and coding the letters was assigned to 
trained coding readers supplied with a carefully prepared manual embody- 


ing detailed definitions of themes and illustrative materials drawn from 
the entries. Checks were made throughout the coding process to insure 
accuracy and consistency of coding. According to the investigators, "sub- 
sequent re-checks have indicated that the coding error is practically neg- 
ligible" (4, p. 26) but no quantitative data on reliability of coding are 
presented. Although all entries were analyzed and coded, provisions were 
made for isolating a random sample of entries for detailed "pilot studies" 
by identifying every tenth entry as the letters went through the coding 
process. Correlations of .98 between total entries and the 10 per cent 
sample in the case of both theme mention and plant participation led 
the investigators to conclude that the sample is adequately representative 
of the entire population. 

In the treatment of MJC entries, frequency of theme incident or men- 
tion has been accepted as providing a quantitative index of employee 
thinking about the job and working conditions, or at least as representing 
the "things that were talked about by employees" in their letters. The 
percentage distribution of theme mention 2 for all divisions of the Gen- 
eral Motors Corporation is shown in Table 18, which presents data from 
the studies conducted independently by the National Industrial Con- 
ference Board and the General Motors Corporation. In order to facili- 
tate detailed comparison of the^two studies, findings have been classified 
by the author under general headings which appear to describe the items 
grouped thereunder. 

It is apparent that the two investigations, involving the use of quite 
different procedures for surveying employee attitudes, yield findings which 
are, in the main, similar in terms of the relative emphasis placed by em- 
ployees upon various aspects of the work situation. In this connection, 
it can, of course, be contended that the two methods were less different 
than was thought to be the case by the General Motors investigators. In 
particular, there seems reason to question the claim that MJC repre- 
sented a nondirective approach to the survey of employee attitudes. The 
advice given by foremen who promoted the contest, the booklets dis- 
tributed to entrants, and the requirement that letters deal only with 
favorable aspects of the job can properly be construed as very definite 
guidance not actually designed "to produce a state of mind that is open 
and undirected" or a climate in which the individual was free to express 
his unconscious desires and reveal his inner personality. On the other 
hand, the absence of formal questions or fixed rating categories un- 
doubtedly introduces a greater degree of freedom for the respondent 
than is present in the more conventional employee-attitude studies. 

Differences in the detailed findings of such studies, as well as the possi- 

3 The coding plan provided for only single coding of the themes mentioned, i.e., 
multiple mentions of a specific theme in any entry were treated as a single mention. 


ble influence of the methods used and of other factors are discussed in 
other chapters. Of interest at the moment is the fact that three studies, 
conducted in different plants, at different times, and using three varieties 
of survey methods, yield consistent results in demonstrating that many 
objects and situations in the working environment promote or interfere 
with the satisfaction of personal needs and wants. In particular, the find- 
ings of employee-attitude surveys confirm the results of experimental 
studies in showing that motivation, as expressed through attitudes and 
sentiments, is far from being solely (or even primarily) a function of basic- 
wages or of wage incentives. These large-scale attitude surveys strikingly 
underline the necessity, referred to in earlier chapters, of focussing atten- 
tion upon the satisfaction of a large variety of needs and wants in a 
planned program for maximizing motivation and achieving optimum 
levels of satisfaction and morale in industry. 

General Motors Corporation 

My Job Contest 

Letter Illustrating Coding Procedure 
Entry No. 56-0994 

Theme Mention 

My Job with General Motor's has given Security 
me what every man wants Security for 
his family I have been with General Mo- 
tor's only five years, a short while compared 
to thousands of other employees. I consider 
it one of the luckiest days of my life when I 

was hired as a timekeeper at the 

Division of General Motors. In less than a 
month I was getting top rate. Since then I 
have been promoted to supervisees clerk 
and stand an excellent chance of becom- 
ing a salaried employee in the near future. 
Knowing that a chance for advancement 
is possible is the best incentive I know of 
for doing a good job. I know from my own 
experience and from others that ability and 
ambition are recognized. More often than 
not a promotion is given to those who have 
shown that they have the ability and the 
ambition for a bigger and better job with 
the Corporation. A Key position with Gen- 
eral Motor's is my goal, by starting at the 
bottom and being able to work yourself to 
a top position is the main reason I like my 
job with General Motors. 

(From C. E. Evans and La V. N. Laseau, My Job Contest (Personnel Psychology 

* t u KT~ 1 \ D^^^^MT,^! Pet/r>Virlrerv Tnr 

Success Theme 

Opportunity for Advancement 



TABLE 18. Findings of Two Employee-Attitude Surveys 

Five Most Important Factors * 

rating by em- 
ployees of 6 



My Job Contest t 

Per cent 
mention of 
58 themes 


Compensation (base pay) 27.9 

- a yj 

Profit-sharing plans (excluding 
employee savings plans) 

Extra-compensation plans 
(all types of bonuses) 

Your burden of state and 
federal taxes 

Job security employment 

Labor turnover rate (its effect 

on morale of employees) 



Benefits from wages 


44.7 Pride in stability of 

2.2 Steady work 

Pride in years of 





Supervisor's temperament 

and personality 
Quality of supervision 
Contact with executives 
Training of supervisors 

and foremen 
Training of executives 



16.3 Management 

10.4 Teamwork 
7.0 Fair treatment 


Opportunity in the com- 
pany for advancement 30.7 

Practice of hiring outsiders 
for responsible jobs vs. 
"promotion from within " 11.1 



Opportunity for ad- 
vancement 25.6 


Training, education, 



Training of wage earners 
( rank-and-file employees ) 

Induction training (includ- 
ing orientation ) 0.8 

* Adapted from Factors Affecting Employee Morale (Studies in 
Personnel Policy No. 85), National Industrial Conference Board, 

t Adapted from Personnel Psychology Monograph No. 1, Per- 
sonnel Psychology Inc., 1950. 


CMC My Job Contest 


NICE Most Important factors 

Per cent 
rating by em- 
ployees of 6 

Per cent 
mention of 
58 themes 


Type of work 18.5 Work type 33.7 

Pride in important 

job 20.2 

Pride in building 

good product 1 .8 


Employee financial benefits 
(group life insurance, sick- 
ness insurance, pensions) 24.4 

Company medical and 

health programs 10.4 

Employees savings and 

thrift plans 5.4 

Insurance 23.7 

Medical facilities 11.5 

Hospital plan 8.9 

Savings plan 7.9 

Pension plans 4.6 


Vacation and holiday 

practices 16.4 

Leave-of-absence practices 

(nonmilitary) 3.3 


Paid vaca tion 15.1 

Paid holidays 2.5 

Leave of absence 1 .8 


Employee feeding facilities 
(restaurants, cafeterias, 
canteens, mobile kitchen) 1 1 .2 

Employee suggestion sys- 
tems 4.1 

Social and recreational ac- 
tivities 3.3 

Recreation 14.0 

Suggestion plan 10.5 

Parties and picnics 6.6 

Cafeteria 5.9 

Parking facilities 2.4 


Practice of informing you of 
your job status (both of 
your success and failures) 19.2 

Employee merit or perform- 
ance rating (an organized 
or systematic method of 
appraising your perform- 
ance) 10.1 

Formal plans for determin- 
ing basic rates of pay ( job 
evaluation programs) 6.1 

Personnel counselling 3.8 

Frequency and method of 

paying wages 3.3 

Employment of mixed races 3.3 


Nondiscrim ina tion 14.3 

Job descriptions 10.1 

Personal history 7.2 

Suitable placement 6.2 

Personnel policies 5.4 
Personnel department 3.4 

Information services 3.2 

Seniority 1.3 

Continued on next page 


TABLE 18 (continued) 

NICE Most Important Factors CMC My Job Contest 

Per cent 

rating by em- Per cent 

ployees of 6 mention of 

companies 58 themes 

Employment methods, in- 
cluding selection, inter- 
viewing and placement 2.9 

Employment record ac- 
curacy and completeness 
of employees' history with 
company 2.6 

Termination allowances 1 .8 

Bulletin boards, house or- 
gans and other methods 
of disseminating informa- 
tion to employees 1 . 5 

Policies and practices regard- 
ing discharge of employees 1 . 1 

Employment of mixed na- 
tionalities 0.9 

Single or married 0.9 

Production drives incentive 

ideas stimulated by company 0.2 


Company's attitude towards Employee relations 5.8 

employee (its interpreta- 
tion of policies whether 
liberal or conservative ) 13.6 

Labor unions 5.8 

Methods of handling griev- 
ances 3.5 

Laws affecting relations 
between employers and 
employees ' 2.6 

Type of union leadership 1.6 

Production drives based 
upon cooperative effort of 
management and employees 1 .0 


Physical working conditions Tools, methods and 

(on the job) 14.4 equipment 16.6 

Total hours worked per day Cleanliness 11.1 

per week 11.7 Working hours 8.5 

Rest periods 7.2 Air and temperature 6.0 

Types and conditions of tools Lighting 5 .4 

and equipment 6.2 Modern plant 3.5 

Shifts (time starting and Washrooms 3.2 

finishing work ) 5 .4 Locker rooms 1 . 1 

Music on the job 5.2 Rest periods 0.8 


NICE Most Important Factors 

Per cent 
rating by em- 
ployees of 6 

Physical conditions of recre- 
ation rooms, lavatories, and 
other personal facilities 4.5 

Length of lunch period 3.6 

Smoking privileges 2.9 

Policy with respect to uni- 
forms (or regulating work- 
ing attire) 1.6 

CMC My Job Contest 

Per cent 
mention of 
58 themes 

Accident-prevention activities 


Company's reputation with 

Type of company product 

( its social importance 

during war) 
Knowledge of company's 

Type of company product 

(its social importance 

during peacetime) 
Company's reputation with 

its customers 
Knowledge of competitors' 

General outside physical 

appearance of plant 
Location of plant ("right 

or wrong side of tracks") 
Your knowledge of com- 
pany's finances 
Knowledge of company's 



5.7 Safety 



Pride in company 32.2 

7.4 Pride in product 22.8 
Comparison with 

other companies 9.4 

6.2 Company and 

America 8.7 

4.9 Free enterprise 8.0 
Pride in commun- 
ity relations 2.6 

4.5 Open house 0.3 



Your own temperament 
ability to get along with 

Your confidence in yourself 

Friendly attitude of all em- 
ployees toward each other 


Comparison other 

Success theme 





Continued on next page 


TABLE 18 (continued) 

NICE Most Important Factors CMC My Job Contest 

Per cent 

rating by em- Per cent 

ployees of 6 mention of 

companies 58 themes 


Transportation facilities (to Plant location and 

and from work) 3.9 transportation 3.5 

Domestic relations and 

home conditions 1 .9 

Housing facilities 1 .5 

Your family's, attitude 

toward your company 

and job 0.1 


1. M. E. Barron, 'The emerging role of public employee counseling,'* Publ. Pers. 
Rev., 1945, 6,9-16. 

2. E. M. Bowler and F. T. Dawson, Counseling Employees. Prentice-Hall, Inc., 1948. 

3. W. J. Dickson, "The Hawthorne plan of personnel counseling/' Amer. J. Orthopsy- 
chiatry, 1945, 15, 343-47. 

4. C. E. Evans and La V. N. Laseau, My Job Contest (Personnel Psychology Mono- 
graph No. 1). Washington: Personnel Psychology Inc., 1950. 

5. Factors Affecting Employee Morale (Studies in Personnel Policy No. 85), New 
York: National Industrial Conference Board, 1947. 

6. J. D. Houser, What People Want From Business. McGraw-Hill Book Co., 1938. 

7. W. H. Kushnick, "A guide to personnel counseling," Personnel, 1943, 20, 139-53. 

8. A. Kornhauser, "Psychological studies of employee attitudes/' /. Consult. Psychol., 

9. McG. Smith, "Mending our weakest links/' Adv. Mgmt., 1942, 7, 77-83. 

10. N. R. F. Maier, Frustration. McGraw-Hill Book Co., 1949, Chapter 8. 

11. E. Mayo, The Human Problems of an Industrial Civilization. Harvard Business 
School, Division of Research, 1946. (First printed by The Macmillan Co., 1933.) 

12. M. L. Putnam, "Improving employee relations," Pers. /,, 1930, 8, 314-25. 

13. F. J. Roethlisberger and W. J, Dickson, Management and the Worker. Harvard 
University Press, 1939. 

14. C. R. Rogers, Client-centered Therapy. Houghton Mifflin Co., 1951. 

15. C. R. Rogers, Counseling and Psychotherapy. Houghton Mifflin Co., 1942. 

16. O. Tead, "Employee counseling: a new personnel assignment its status and its 
standards," Adv. Mgmt. 9 1943, 8, 97-103. 

17. R. S. Uhrbrock, "Attitudes of 4430 employees," J. Soc. Psychol., 1934, 5, 365-77. 

18. G. Watson, "Work satisfaction" in G. W. Hartmann and T. Newcomb (Eds.), 
Industrial Conflict. The Cordon Co., 1939, 114-24. 

19. T. N. Whitehead, The Industrial Worker. Harvard University Press, 1938. 

20. J. Wilensky and J. Wilensky, "Personnel counseling: The Hawthorne Case," 
Amer. J. SocioL, 1951, 57, 265-80. 


The Assessment of Employee 

COMPANIES which conduct employee-attitude surveys are 
almost invariably interested in obtaining some kind of an over-all score 
which gives an impression of the employee's general "attitude toward the 
company, its policies, and its treatment of its personnel" (13). 

The Computation of a Morale Index 

from an Interview Survey and Its Use 

in Departmental Comparisons of Morale 

A frequent practice in industry has been to arrive at such an over-all 
score, generally known as a morale index, from either all or some of the 
questions included in the survey or from special scales used in conjunc- 
tion with questionnaires (23a). An investigation by Houser, a pioneer 
in employee-attitude surveying, illustrates the use of data from a highly 
patterned interview in arriving at an over-all morale score and in com- 
paring levels of employee morale in various departments of a company 
(21). In this study, conducted in the 1920's, nineteen features of the 
work situation, subsumed under the captions of ( 1 ) Adjustment to the 
Job; (2) Supervision; (3) Incentives; (4) Participation-Expression; and 
(5) General Working Conditions and Facilities, were selected as repre- 
senting the most important elements in determining the relationship 
between employees and the organization. 

A random sample of employees in the organization was then inter- 
viewed individually by a skilled interviewer an outsider without knowl- 
edge or preconceived ideas about the organization. The interview was 
conducted, in this case, by means of carefully standardized questions, 
put always in the same way with the same inflections. The following ques- 



tions, asked in connection with "adding to ability/' considered to be one 
aspect of Incentives, are characteristic of the type used in this investiga- 
tion. "Are you learning things on your job that mil be of use to you later 
on, either on this job or on a higher one? Are you getting a chance to learn 
and study some other job? How much do you feel that you are growing 
on the job?" (21, p. 177). 

The employee's answer was graded by comparison with a rating scale 
consisting of five "type" responses, separated by approximately equal 
intervals from the viewpoint of degree of feeling. This scale had been 
memorized by the interviewer so that there was no need of referring to 
it in the course of the conference. The following "type" responses, used 
in grading the answers to the questions on "adding to ability," give an 
indication of the nature of the scoring criteria. 

"5. The company certainly does encourage me and offers me every op- 
portunity to develop and make progress. I'm sure I am being given every 
chance I could be. 

"4. Yes, there are a number of pretty good chances ahead. I believe 
Tm getting some new knowledge and ability every day. It could be a little 
better, perhaps, but the company takes an interest and that's an encour- 

"3. The chance for learning here is all right, I guess about average 
no kick as far as I can see. 

"2. You've got to pick most of it up for yourself. The job doesn't give 
you much chance. I've learned about all I can. I feel that more interest 
should be shown along this line. 

"1. Don't think I'm getting along at all! I'm in a fierce rut! No chance 
to learn! There's no encouragement at all to try to learn or go ahead" 
(21, p. 178). 

Each response was assigned a numerical value ranging from "1," lowest 
possible score indicating unqualified hostility, through "3," the middle 
point, indicating neutrality or indifference, to "5," the highest possible 
score indicating unqualified enthusiasm or favor. The average of the 
scores on responses to all the questions was taken to represent the over- 
all level of morale of the individual employee. By averaging the scores 
for employees in individual departments it becomes possible, as shown 
in Figure 21, to compare the over-all levels of morale characterizing groups 
of employees in different departments. 

It is apparent, from Figure 21, that there can be wide differences in 
morale among different departments of the same company. In fact, as 
shown in a study of four organizations reported by Hull (23), variations 
in the level of employee morale from department to department within 

Scale 10 

Avtrageof Departments 
( Company Index) 





Department 1 

Department 2 

Department 3 

Department 4 


Midpoint of Indifference 
or Neutral Fueling 

Figure 21. Departmental Variations in Employee Morale* 

a single company can be far greater than the differences in average morale 
scores among the companies. Determination of the sources of such dif- 
ferences lays the foundation for constructive action directed towards the 
improvement of the personnel situation in departments characterized by 
low morale scores, as well as in the company as a whole. 

Computation of a Morale Index from an 

Attitude Scale and Its Use in Comparing 

Occupational Groups 

Morale scores may be used not only for the comparison of levels of 
morale in various departments and companies, but to investigate the 
morale of various segments of the employee population. Uhrbrock (49) 
furnishes a classic example of utilization of an employee-attitude survey 
for this purpose. In this case, use was made of an attitude scale constructed 
in accordance with principles formulated by Thurstone (46, 47) to ob- 
tain the basic data for appraising employee morale. Fifty statements with 
scale values from 0.6 to 10.5, including items pertaining to varied com- 
pany policies and practices, were included in the final employee attitude 
scale which, upon analysis, was shown to have a reliability (odd-even) 
of .891 : .007 (corrected by the Spearman-Brown prophecy formula for 
length). The following illustrate both the types of statements used and 
their scale values 

* By permission, from J. D, Houscr, What the Employer Thinks, Harvard University 
Press, 1927. 



Scale value 

10.4 I think this company treats its employees better than any 

other company does. 
8.9 A man can get ahead in this company if he tries. 

The company is sincere in wanting to know what its em- 
ployees think about it. 

I believe accidents will happen no matter what you do about 

The workers put as much over on the company as the company 
puts over on them. 
Soldiering on the job is increasing. 
I believe many good suggestions are killed by the bosses. 
I think the company goes outside to fill good jobs instead of 
promoting men who are there. 
In the long run this company will "put it over' 9 on you. 





The attitude scale, with the statements arranged in random order in 
so far as scale values are concerned, was presented to 3,934 factory em- 
ployees, 96 clerical workers and 400 foremen in various plants of the 
company, with appropriate instructions, including the request that the 
blanks be returned without signatures. 


M4 20 
















Figure 22. Morale Scores of Factory Workers, Clerical Workers, and Fore- 
men * 

The average scores of each of the three groups of employees and the 
total distribution of scores are shown in Figure 22. Differences among 
the average scores of the three groups are statistically significant. It is 
apparent that the average score of factory workers (6.34) is approximately 
at the midpoint of the scale (6.0) and that the distribution is approxi- 
mately normal. In contrast, the average score of clerical workers (6.89) 

* By permission, from R. S. Uhrbrock, "Attitudes of 4,430 employees/* J. Soc. 
Psychol., 1934, 5, 365-77. 


and the skewed distribution of measures shows a considerably more favor- 
able attitude towards the plant than exists among factory workers. Even 
more favorable is the attitude of foremen with an average score of 7.19 
with few scores below 5, and a considerable number of scores at the higher 
end of the scale. Further analyses also showed slight, although in some in- 
stances statistically significant, differences among the various factories 
operated by the company, with average morale scores for factory worker 
groups ranging from 5.97 to 6.76, and in the case of foremen from 6.49 
to 7.59. 

The Relation between Specific Factors and 
the Morale Index 

The study by Houser (21), described on pages 271-273, provides an 
illustration of a morale index computed from responses given to questions 
asked during an interview. There are also many instances where the morale 
index is computed by totaling scores on all or a selection of items in a 
written questionnaire. One example of this is to be found in the Florida 
Power and Light Company survey (43), described on pages 325-326. 
Another illustration appears in surveys conducted by Arthur Kolstad As- 
sociates, in which a morale score is computed separately for each employee 
on the basis of answers to ten general attitude questions included, with 
others, in the survey questionnaire. In this approach, the employee who 
has a morale score of 100 the maximum on this scale is one who 

"1. Believes his company to be one of the very best of all companies 
as a place to work. 

"2, Thinks the management of his company care more about the wel- 
fare of its employees than does the management of any other company. 

"3. Knows of no other company where he would rather work if he 
could get an equally good job elsewhere. 

"4. Blames himself and not the company for any dissatisfaction he may 
have experienced on the job. 

"5. Thinks the company does much more than one might expect to 
promote good working relationships between himself and the people with 
whom he works. 

"6. Feels that he is really regarded as a part of the organization. 

"7. Thinks the management always fair with people in jobs such as his. 

"8. Thinks the people just above him are always fair with him. 

"9. Knows of no other company that treats its employees as well as 
his does. 

44 10. Believes that he can be sure of his job as long as he does good 
work" (23). 

In some instances, simultaneous use is made of a number of survey 
tools, including frequently an (tttitude scale, such as described on page 
273, a questionnaire, and a data sheet for obtaining information on the 


employee's age, sex, length of service, department, job status, etc. In 
such cases, the morale index is generally derived from the attitude scale, 
which may also be broken down to furnish a distribution of responses 
to specific items included in the scale. The questionnaire provides data 
on employee attitude toward specific factors in the plant situation, par- 
ticular aspects of the industrial-relations program, etc. Illustrative analyses 
of questionnaire items, taken from a report by Bergen, are presented in 
Table 19. Through such analyses, according to Bergen, "the efficacy of 
various practices is established and the need for revision in present prac- 
tices or installation of new practices determined" (6, p. 55). The factual 
data are employed in stratifying the working force in relation to depart- 
ment, job, age, sex, etc., for investigating differences in morale and in at- 
titudes toward specific policies and practices. 

The use of multiple devices most particularly where the measure of 
morale and factual data are obtained as independent criteria provide 
an opportunity for a series of significant comparisons. In particular, it 
becomes possible to relate the degree of satisfaction or dissatisfaction 
with specific conditions and policies to morale. In general, although a 
variety of statistical methods have been used, this involves a determina- 

TABLE 19. Results of a Multiple-Choice Questionnaire (after Bergen) 


Question: "When business gets really bad, 'which of the following steps 
do you think the company should take? (Check the one you think should 
be done first.)" 

Per cent 

Cut working hours and spread the work to everybody 69.7 
Lay off some people and thus give more hours to those 

who remain 

Reduce salaries and wages 
No answer 



Question: "When business gets really bad, which of the following steps 
do you think the company should take? (Check the order in which you believe 
the steps should be taken 1, 2, 3, 4, 5, 6.)" 


1. Reduce or cut out dividends 1.30 

2. Cut advertising and sales promotion expense 3.15 

3. Reduce salaries 3.29 

4. Reduce factory workers' wages 3.66 

5. Reduce forces 3.92 

6. Take losses and risk going out of business 5.25 


tion of the percentage of employees with high and low morale, respec- 
tively, who express satisfaction or dissatisfaction with specific factors in 
the plant situation covered in the questionnaire. It thus ''becomes possi- 
ble to rank the specific factors in the company's personnel program in 
the order of their significance as determinants of general morale" (6, p. 56). 
As indicated in later sections of this chapter, caution must be exercised 
in the evaluation of such relationships. Nevertheless, as will also appear in 
later chapters, information of considerable value in the assessment of over- 
all attitudes has been derived from such comparisons. 



The analysis of employees' attitudes and morale frequently includes 
breakdowns with respect to age and length of service. In fact, a survey 
conducted by the National Industrial Conference Board (13), in 1951, 
shows that such breakdowns are second in frequency only to breakdowns 
by departments. Illustrative of the analysis of age and length of service in 
relation to morale is a study by Benge and Coppell (5). In this, a ques- 
tionnaire, designated an "opinion ballot/' was used to obtain the reac- 
tions of 722 employees in five widely scattered plants of a confectionary 
manufacturing company (a) towards 28 items pertaining to the em- 
ployee's job and (b) towards 20 topics relating to company policies, 
activities and personnel services. Replies to the 28 items were used in the 
computation of morale indices, weighted in accordance with the number 
of responses to each item. 

Such indices were calculated for each of the 28 items when employees 
were classified by age and length of service as well as on the basis of de- 
partments, shifts, etc. Indices for each of these classifications were then 
combined to obtain composite attitudes toward job, toward boss, toward 
company. Finally, over-all or average morale indices were computed and 
comparison made between age groups, length-of-service groups, depart- 
ments, plants, etc. 

As appears in Figure 23, the comparison of age groups showed that, 
in the five plants under consideration, employees under 20 years of age 
had relatively high morale; those between 20 and 29 years of age, rela- 
tively low morale, and that morale rises again with each successive group- 
ing beyond 30 years of age. Similar results appeared in the comparison 
of length-of-service groups. As shown in Figure 24, morale was found 
to be highest among employees with less than two years of service; the 
level of morale fell consistently with increasing length of service, with 
a slight upturn occurring among employees with ten and more years of 






30*39 4O49 &0-59 

Figure 23. Morale According to Age * 

Progressive deterioration in employee morale following the first few 
years of service has also been observed in other surveys. In a study in a 
manufacturing company, by Bergen, male employees with two to seven 
years of service were found to have lower average morale scores than those 
with shorter or longer service (6) . Findings from plant-wide samplings of 
rank-and-file manufacturing employees, conducted by the Industrial Re- 
lations Center, University of Minnesota, showed a statistically significant 
relationship between length of service and over-all attitudes toward the 
company, in that employees on a job five years or more have higher aver- 
age attitude scores than those on a job from one to four years (48) . These 
findings may appear to be inconsistent with those cited above, but data 
are not broken down in a manner to make possible exact comparison with 
other studies. In the same investigation, employees 45 years of age and 
over were found to have higher average attitude scores than did employees 
30-44, and 20-29 years of age, respectively. 

Hull has suggested that the decrease in morale following the first few 
years of service "is probably a process of disillusionment involved in the 
average worker's adjustment to the job/' According to this explanation, 
the worker is filled with enthusiasm at first, the job is more interesting 

* By permission, from E. J. Benge and D. F. Copell, "Employee morale survey," 
Modern Mgmt., (January) 1947, 7, 19-2L 





UNOEfc 1-2 

3-4 5-9 

KM9 20 


Figure 24. Morale According to Length of Service * 

because it is new and his hopes may be high. "Perhaps this sense of 
achievement is lessened after the early period of easy progress is past; cer- 
tainly there is no opportunity for all ambitious or hopeful newcomers 
to get ahead as fast as they may at first hope/' On the other hand, ac- 
cording to Hull, "employees who remain on the payroll for more than five 
or ten years are a selected group, having survived the constant process 
of elimination through dismissal or resignation. If they are of outstand- 
ing ability that ability may have been recognized. Those who have not 
advanced have accepted the fact that there is not room at the top for all 
and have perhaps resigned themselves to their fate, a fate which does not 
seem so bad after all" (23, p. 168). 

Support for this viewpoint is suggested in the fact that in the six-plant 
1947 National Industrial Conference Board survey (14), described in 
Chapter 13, "opportunity for advancement" was rated as the most impor- 
tant factor by very few employees with less than six months of service 
and by not a single one of the 20-year-and-over employees. Similarly, this 
factor appeared very infrequently among the five most important factors 
as rated by newly hired and longer-service employees. 

As suggested on page 236, comparisons between age and length of serv- 
ice groups suffer from the fact that consideration is generally not given 
to possible differences among the jobs held by the various subgroups. 

* By permission, from E. J. Benge and D. F. Copell, "Employee morale survey," 
Modern Mgmt., (January) 1947, 7, 19-22. 



Furthermore, in most instances, no attempt is made to determine the 
relative influence of age and length of service upon morale. Since em- 
ployees grow older with increasing length of service, it is to be expected 
that the same trends will appear with increasing age as are observed with 
increasing length of service. Comparison of Figures 23 and 24 suggests 
that age is a more potent factor than length of service in producing the 
increment of morale which has been observed among workers with con- 
siderable length of service. This possibility appears particularly in find- 
ings reported by Hull which show, as indicated in Figure 25, that in spite 
of the deterioration of over-all morale with increasing length of service, 
average morale scores moved upward with age within separate length-of- 
service brackets (23), 



21 TO 25 YEARS 


t 21 YEARS 

2 1 TO 25 YEARS 
26 TO 30 YEARS 
3 1 TO 40 YEARS 



21 TO 25 YCARS 
26 TO 30 YEARS 
31 TO 40 YEARS 



21 TO 25 YEARS 
26 TO 30 YEARS 
31 TO 40 YEARS 


Figure 25. Morale Scores by Age Groups with Length of Service Held Con- 
stant * 

Such findings may be of particular significance in considering the em- 
ployment of older people, since they suggest that older applicants do not 
present problems in so far as attitudes towards the company and the job 
are concerned. On the other hand, the data bearing on the relation be- 

* By permission, from R. L. Hull, "Measuring employee attitudes a proving 
ground for personnel policy and practices/' Conference Board Mgmt. Record, (Novem- 
ber) 1959, 165-72. 


tween length of service and morale suggest that management cannot 
necessarily depend upon the satisfaction and loyalty of employees with 
long years of service in itself to carry the plant through periods of poten- 
tial or actual labor strife. Of interest in the latter connection is the fact 
that of three age groups subjected to analysis in the 1947 National In- 
dustrial Conference Board study (14), a significantly larger percentage 
of oldest employees (50-years-and-over group) indicated a conviction that 
labor unions have an important effect on their attitudes towards their 
job and company, in spite of the further fact that less than one per cent 
of all employees in the six companies studied gave this factor top rank- 
ing (Table 17). 

Problems in the Interpretation and 
Evaluation of Morale Indices 

From the studies described in earlier sections, it is apparent that a 
well conceived index can serve useful purposes in giving an impression 
of over-all morale and in making meaningful comparisons within the 
plant. Such an over-all score also furnishes a practical medium for arriv- 
ing at broader generalizations concerning employee morale in American 
industry when a standard method for computing the morale index is 
used in a sufficiently large number of companies. However, in the treat- 
ment of morale indices, as of other aspects of attitude surveying, there 
are instances of oversimplification and overgeneralization, and of other 
weaknesses and deficiencies in dealing with the data. These appear par- 
ticularly in analyses of relationships between the morale index, on the 
one hand, and specific attitudes, on the other. 


As noted earlier, the purpose of such comparisons is to identify the 
favorable and unfavorable attitudes with respect to the work situation 
which account for high and Zow morale, respectively. The usual assump- 
tion in such analyses is that a specific factor is a significant influence 
in determining low morale when employees with low morale scores more 
frequently express unfavorable attitudes toward the factor in question 
than do workers with high morale scores. Similarly, it is assumed that a 
specific factor is a determinant of high morale if a large proportion of 
workers with high morale scores exhibit favorable attitudes toward the 
factor. If the difference in percentages between such workers in attitudes 
toward the factor is slight, it is looked upon as having little or no influ- 
ence as a determinant of morale level (23, 37). 

The use of this method without reference to other figures can intro- 
duce serious errors into the evaluation of the morale situation in a plant. 


The character of such errors is illustrated by Kornhauser with reference 
to an investigation, by Houser (22), in which it was reported that 93 
per cent of employee responses revealed a feeling in favor of higher pay. 
However, in ranking the relative importance of job factors, pay appeared 
almost at the foot of a long list of such factors since, when the "method 
of differential comparison" described above was applied, employees with 
high morale scores were almost as frequently in favor of higher pay as 
those with low morale scores. "Clearly/* states Kornhauser, "all that this 
last conclusion can mean is that differences in attitude about pay were 
not important in accounting for differences of morale within that com- 
pany. But what of morale in that company compared with other com- 
panies? Surely management is led to a disastrously complacent conclusion 
when it is told that even though 93 per cent of its employees think that 
they should have more pay this is a relatively unimportant influence 
upon their morale. The procedure is in danger of arriving at similarly mis- 
leading conclusions wherever it is employed" (37, p. 138). 

Other considerations, of the kind discussed in Chapter 12, enter into 
the evaluation of conclusions from survey findings concerning morale and 
its determinants. Here, as in other comparisons, there is a tendency to 
ascribe causal relations to a factor without considering the highly intricate 
interrelationships among all the^ factors which are identified as having 
significance in determining the level of morale. Also, in some instances, 
employees are asked to give direct ratings concerning the relative sig- 
nificance of a list of work characteristics as a means of arriving, among 
other things, at the importance of each in determining over-all attitudes. 
As pointed out in the discussion of the 1947 National Industrial Con- 
ference Board survey described in Chapter 13, employees are in such 
cases required to make many implicit assumptions in defining the char- 
acteristics and also concerning the range of variation within any of the 
items which are being ranked. Interpretation of findings must take into 
account the influence of differences in such assumptions upon the rank- 
ing of factors (37). Nevertheless, in spite of these and other hazards of 
interpretation, comparisons within the company can be very valuable. 
Furthermore, comparison of results bearing upon morale obtained from 
a variety of company surveys represents a useful step in arriving at gen- 
eralizations concerning the potency of particular incentives and deter- 
minants of morale, and in formulating inferences concerning motives 
underlying behavior in industry. 

Conceptual and Operational Definitions 
of Morale 

As indicated on page 240, a disturbing factor in arriving at broad gen- 
eralizations concerning morale is the almost consistent failure of surveys 


(and also of experimental studies) to deal with the problem within an 
appropriate context of theory. As noted, the most usual tendency has 
been to define the dimensions of morale in terms of what is revealed by 
the investigation. Another trend has been to leave the interpretation 
of this term to the subjects of the investigation. In general, there has been 
considerable neglect of the dictum, as expressed by McNemar, that "two 
jobs need to be done: the determination of the dimensions of morale 
and the construction and validation of scales for measuring these dimen- 
sions" (39a, p. 365). In only relatively few instances, as in studies by 
the Survey Research Center (29, 30), has the design of the study been 
developed around clearly stated operational or conceptual definitions of 
morale. There appears to be urgent need for the wider use of sound 
theoretical concepts in planning investigations of morale and of its de- 

The nature of available conceptual definitions, as derived from per- 
sonality theory and social theory in psychology, is illustrated in three 
definitions, which take "cognizance of three different realms of discourse/' 
resulting from a Conference on Psychological Factors in Morale held, 
in 1940, under the auspices of the Division of Anthropology and Psy- 
chology, National Research Council. These, described by Child as use- 
ful guides to further work, are as follows 

44 1. (The individual-organic emphasis). The term morale refers to a 
condition of physical and emotional well-being in the individual that 
makes it possible for him to work and live hopefully and effectively, feel- 
ing that he shares the basic purposes of the groups of which he is a 
member; and that makes it possible for him to perform his tasks with 
energy, enthusiasm, and self-discipline, sustained by a conviction that, 
in spite of obstacles and conflict, his personal and social ideals are worth 

"2. (The group emphasis) . Morale refers to the condition of a group 
where there are clear and fixed group goals (purposes) that are felt to be 
important and integrated with individual goals; where there is confi- 
dence in the attainment of these goals; and subordinately, confidence 
in the means of attainment, in the leader, associates, and finally in 
oneself; where group actions are integrated and cooperative; and where 
aggression and hostility are expressed against the forces frustrating the 
group rather than toward other individuals within the group. 

"3. (Emphasis on individual-within-the-group on any specific occa- 
sion). Given a certain task to be accomplished by the group, morale per- 
tains to all factors in the individual's lire that bring about a hopeful and 
energetic participation on his part so that his efforts enhance the effec- 
tiveness of the group in accomplishing the task in hand" (12, p, 393). 

Similar conceptualizations of morale are found in definitions formu- 
lated by other psychologists, including Allport (1, p. 4), Watson (50, p. 
275, 51), Lippitt (38, p. 120), and others. These are generally derived 


from realms of discourse considered by the ten psychologists involved in 
the National Research Council conference on morale and, in general, 
there is considerable overlap among the definitions. Other definitions 
are more specifically oriented from the operational point of view (7, 19). 
So, according to Blankenship, 

"As the term morale is used ordinarily by the employer, laborer, and 
psychologist alike, it refers to a feeling or 'togetherness/ There is a sense 
of identification with and interest in the elements of one's job, working 
conditions, fellow workers, supervisors, employers, and the company. The 
more a worker possesses such feelings, the higher his morale" (7, p. 299) . 

There are common threads in the definitions of morale presented by 
various social scientists. Included in a number is the concept of striving 
toward the attainment of future, as contrasted with immediate goals. 
Reference is found to release from tension, or the attainment of physical 
and mental well-being, but again with respect to long-term rather than 
to immediate adjustment. Most common to both conceptual and opera- 
tional definitions is the emphasis upon group aspirations and the integra- 
tion of individual goals with those characterizing the group. This char- 
acterization of morale has been well summarized by French in the state- 
ment that "The abundant literature on morale contains many and varied 
definitions of the term, but in those definitions which imply any opera- 
tions for observing morale (and the number of such is considerably less 
than the total) there seems to be tacit agreement that the term denotes 
aspects of group behavior indicative of group effectiveness ... If it can 
be shown that groups which achieve their goals efficiently exhibit a high 
degree of cohesiveness, think well of their leaders, do not fight much 
among themselves, agree on their objectives, have confidence in their 
equipment, and so on, then these manifestations represent high morale; 
but only if a relationship to goal attainment can be shown'' ( 16, p. 465) . 

Such emphasis upon the group is found also in the succinct definition 
given in Chapter 1, viz.: "Morale is an attitude of satisfaction with, de- 
sire to continue in, and willingness to strive for the goals of a particular 
group or organization" (44, p. 1). Acceptance of this viewpoint leads 
immediately to a highly significant problem in research, namely that 
of identifying the group (from the many in the industrial plant with 
which the individual is associated) which may exercise the most potent 
influence in the development of high morale levels (3, 8, 9, 17). Thus, 
studies of the American soldier during World War II, reported by 
Stouffer et al. (45), appear to indicate that in the case of the combat 
soldier, identification and loyalty to the Army as a whole was far over- 
shadowed by his identification and loyalty to his own primary group, i.e. 
to the smaller unit of the Army to which he was attached. There is 


considerable evidence from studies cited in other chapters of this volume 
that in the industrial scene morale may be directed toward the small 
work group characterized by goals which can be in conflict with those of 
the larger social organization, the company. This problem has been con- 
sidered in earlier chapters x and is also dealt with further in Chapter 23. 
As is apparent from the consideration of definitions of morale in pre- 
ceding pages, diverse theoretical positions may nevertheless be taken in 
undertaking research on morale. The important need, in the design of 
studies, is for some commitment in the way of "psychological formula- 
tions which will serve to organize thought and inquiry" and provide 
guideposts for broad generalizations concerning the determinants and 
significance of morale. In this, as in other areas of attitude surveying, "the 
development and improvement of these interpretative constructions, 
through industrial studies as well as by all other means, may prove to be 
a most valuable contribution psychologists can make in this field" (37, p. 



In the investigations discussed above and in many others, as illustrated 
in the work of Kolstad (36), Kerr (31, 34), Harris (20), Katz (25), and 
others, interview, questionnaire, or attitude-scale items have been com- 
bined to yield a single over-all index of employee morale. Attention has 
also been directed towards the possibility of using elements of individual 
work performance or plant behavior to provide an objective index of em- 
ployee morale. 

Turnover and Absenteeism: According to French, evidence of a dy- 
namic sort concerning morale might be obtained by observing group 
reactions in standardized situations (16). Suggestive possibilities here, 
he adds, are false news announcements (40) and various situations of 
a more conventional type such as intergroup competitions, standard group 
problems, etc. Among objective indices which have received considerable 
attention are absenteeism 2 and turnover. For example, investigations by 
Mayo and his associates in metal-working plants and in the aircraft in- 
dustry, described in Chapter 11, were based on the assumption that ab- 
senteeism and turnover rates represent indices of involvement in achiev- 

1 See particularly pages 176-178; 206-216. 

2 Evidence that absenteeism may be a form of group behavior is found in a study, by 
Walker, which showed that a distribution of absences in industry assumes the form 
anticipated from the application of Allport's J-curve hypothesis of conforming be- 
havior. Sec K. F. Walker, 'The application of the J-curve hypothesis of conforming 
behavior to industrial absenteeism," /. Soc. Psychol., 1947, 25, 207-16. F. H. Allport, 
"The J-curve hypothesis of conforming behavior/ 1 /. Soc. Psychol., 1934, 5, 141-83. 


ing the objectives of the work group (15, 39). Consideration of findings 
from studies in industry has led Young to conclude that absenteeism and 
turnover are "symptoms of and causes of unsatisfactory working con- 
ditions, poor supervisory-employee relations and relations between em- 
ployees" (52). According to Baldamus and Behrens, comparison of ab- 
senteeism for different days of the week can yield an objective measure 
of employee morale. The underlying assumption here is that "owing to 
cumulative fatigue, the number of absentees in a factory should be ex- 
pected to increase from Monday to Friday; if in reality the opposite 
happens, factors other than fatigue or working conditions (which do 
not change from day to day) must be involved, notably morale" (4, p. 

Both absenteeism and turnover have advantages as potential indices 
of morale because of the objectivity and accuracy with which such events 
can be recorded. Furthermore, studies conducted in South Africa, by 
the National Institute for Personnel Research, present the interesting 
possibility of identifying "absent prone" individuals and of predicting 
from a year's record who will be the "absence offenders" in the succeed- 
ing years (2). This would furnish the possibility of taking preventive ac- 
tion with respect to the individual worker which, under the assumptions 
stated above, would constitute action toward the improvement of morale. 
However, studies by Kerr (32, 33, 35) have suggested that different types 
of absenteeism, e.g., total, excused, unexcused, vacation, are unrelated 
and that absenteeism is far from being a unitary trait. This conclusion 
is derived from intercorrelations of diverse variables, such as departmental 
average job satisfaction; number of employees in department; various 
types of absenteeism; degree of incentive work; labor turnover, etc. Fur- 
thermore, "triple audit" techniques employed by the Industrial Rela- 
tions Center, University of Minnesota, have shown, in a small number 
of manufacturing plants, that while "employees with no absenteeism in 
the past 6 weeks have higher average attitude scores than those absent 
3 or more times," the relationship is not statistically significant (48, p. 46). 
Furthermore, findings from the South African study, referred to above, 
have suggested that the tendency toward frequent absences, i.e. "absence 
proneness," is, at least in part, "person-centered" (2). In other words, 
the indications are that absenteeism is a function of health, aptitude 
for the job and social background rather than exclusively (or perhaps 
even in large part) an index of morale. 

Efficiency as an Index of Morale: Such facts suggest that much more 
is needed in the way of research before an objective record of absentee- 
ism can be substituted for morale indices derived from interview, ques- 
tionnaire and attitude-scale items. Another approach to the prediction 


of employee morale from objective data is to be found in the work of 
Giese and Ruter (18). Underlying this is the viewpoint that although 
the use of correctly designed questionnaires has produced valuable (and 
even startling) results, the questionnaire method is "cumbersome, costly, 
and slow" and fails to produce findings which are cost accountable. A 
combination of six objective performance measures, including per cent 
productive efficiency, per cent error efficiency NOT affecting customers, per 
cent errors efficiency affecting customers, per cent turnover, per cent late, 
and per cent absent yielded a correlation of .71 with departmental morale 
scores obtained through the use of a conventional questionnaire in a re- 
tailing organization. From such data, the conclusion is drawn that the 
use of such objective factors for morale measurement "is not only practi- 
cal, but meaningful to executives in business and industry/' 

It is to be noted that (a) this approach involved the use of a "morale 
questionnaire" as the criterion; (b) it orients the problem of morale as 
being primarily associated with the objective of reducing costs; (c) both 
the factors and relative weightings are probably specific to the particular 
business. Nevertheless, the study suggests possibilities in the use of work 
performance factors for the assessment of morale. 


During recent years, investigators have undertaken the investigation 
of the dimensions of morale through the use of factor analysis techniques. 
This approach has been taken in both military (44) and industrial situa- 
tions (24, 26). Preliminary reports of such research by the Survey Re- 
search Center, University of Michigan, have suggested that at least four 
factors are necessary to account for observed interrelations among morale 
items, namely, job satisfaction; mobility or progress; supervision, and 
satisfaction with the company (28, 42). These factors were selected by 
factor-analyzing a morale questionnaire using the multiple group method. 
However, according to Schooler, if the criteria for rotation be altered, "it 
is conceivable that the total configuration could be explained by factors 
that have meaningful labels and are not the same as those already 
discussed" (42). Schooler also notes that when the centroid method is 
applied to the same questionnaire data, changes occur in the factors 
extracted. In particular, the mobility factor does not appear. Never- 
theless, the results show that one factor alone does not sufficiently ac- 
count for the correlational matrix. Findings from this study do not sup- 
port the view that there is a unitary or general "morale factor" and 
suggest, on the contrary, that morale "is in fact a complex of several 
kinds of attitudes." 

Research, involving the application of factor analysis methods by the 


Employee Attitude Research Group of the Industrial Relations Center 
at the University of Chicago has also suggested "... that employee at- 
titudes and morale are a complex a configuration of many forces and 
variables" (10). Such research, associated with the development of the 
SRA Employee Inventory, has led to the tentative identification of 
seven independent factors or dimensions of morale, viz., (1) Personal 
Rewards; (2) Immediate Supervision; (3) Company Operations; (4) 
Psycho-physical Conditions of Work; (5) Job Satisfaction; (6) Work 
Relations; and (7) Integration in the Organization. 

In discussing progress in the factorial analysis of the dimensions of 
morale, Burns carefully notes the tentative character of such classifica- 
tions and the need for further research in this area. Similarly, much 
is still needed in the >vay of additional research on "experiential" (11) 
and other objective or "behavioral indices" of morale (41, 44) . In particu- 
lar, considerable work remains to be done in establishing the reliability 
and validity of both these and more commonly used morale measures. 
In general, although studies of more objective measures are under way, 
the fact remains that the combination of specific items embodied in 
questionnaires and attitude-scales, and data derived from interviews, of 
the types cited in this chapter, have so far yielded the most productive 
information concerning levels of employee morale in industry and busi- 


1. G. W. Allport, "The nature of democratic morale" in G. Watson (Ed.), Civilian 
Morale. Houghton Mifflin Co., 1942, 3-18. 

2. A. G. Arbous and H. S. Sichel, New Techniques for the Analysis of Absenteeism 
Data, South African Council for Scientific and Industrial Research, National In- 
stitute for Personnel Research, 1952 (Unpublished study). 

3. E. W. Bakke, Adaptive Human Behavior. Yale University Press, 1950. 

4. R. Baldamus and H. Behrens, "Variations in absenteeism during the week: an 
index of employee morale," Nature, 1950, 165, 831-32. 

5. E. I. Benge and D. F. Copell, "Employee morale survey/' Modern Mgmt., (Jan- 
uary) 1947, 7, 19-22. 

6. A. B. Bergen, "Finding out what employees are thinking/' Conference Board 
Mgmt. Record, (April) 1939, 53-58. 

7. A. B. Blankenship, "Methods of measuring industrial morale" in G. W. Hartmann 
and T. Newcomb (Eds.), Industrial Conflict. The Cordon Co., 1939, 299-312. 

8. H. Blumer, "Morale" in W. F. Ogburn (Ed.), American Society in War Time. 
University of Chicago Press, 1943, 207-32. 

9. H. Blumer, "Collective behavior, Part IV" in A. M. Lee (Ed.), New Outlines of 
the Principle* of Sociology. Barnes and Noble, Inc., 1946, 205 ff. 


10 R. Burns, "Employee morale its meaning and measurement/' Proceedings 
Fourth Annual Meeting, Industrial Relations Research Assn., Boston, Mass.: 
1951, 52-68. 

11. R. B. Cattcll, 'The ergic theory of attitude and sentiment measurement," Educ. 
Psychol Meas., 1947, 7, 221-46. 

12. I. L. Child, "Morale: A bibliographical review/' Psycho/. Bull, 1941, 38, 393- 


13. Experience with Employee Attitude Surveys (Studies in Personnel Policy No. 
115). New York: National Industrial Conference Board, Inc., 1951. 

14. Factors Affecting Employee Morale (Studies in Personnel Policy No. 85). New 
York: National Industrial Conference Board, Inc., 1947. 

1 5. J. B. Fox and J. F. Scott, Absenteeism: Management's Problem (Business Research 
Studies No. 29). Harvard University, Graduate School of Business, Division of 
Research, 1943. 

16. R. L. French, "Morale and leadership" in Human Factors in Undersea Warfare. 
Washington: National Research Council, Committee on Undersea Warfare, 
1949, 463-90. 

17. B. B. Gardner, "An approach to the problems of morale" in Personnel Administra- 
tion in Libraries. University of Chicago, Graduate Library School, Library Insti- 
tute, 1946, 85-93. 

18. W. J. Giese and H. W. Ruter, "An objective analysis of morale," /. Appl. Psychol., 

19. J. Harding, "The measurement of civilian morale" in H. Cantril (Ed.), Gauging 
Public Opinion. Princeton University Press, 1944, 233-60. 

20. F. J. Harris, "The quantification of an industrial employee survey: I, Methods: 
II, Application," /. Appl. Psychol, 1949, 33, 103-13. 

21. J. D. Houser, What the Employer Thinks. Harvard University Press, 1927. 

22. J. D. Houser, What People Want from Business. McGraw-Hill Book Co., 1938. 

23. R. L. Hull, "Measuring employee attitudes a proving ground for personnel pol- 
icy and practices," Conference Board Mgmt. Record, (November) 1939, 165-72. 

23a. R. L. Hull and A. L. Kolstad, "Morale on the job" in G. Watson (Ed.), Civilian 
Morale. Houghton Mifflin Co., 1942, 349-64. 

24. R. L. Kahn, "An analysis of supervisory practices and components of morale" in 
H. Guetzkow (Ed.), Groups, Leadership, and Men. Pittsburgh: Carnegie Press, 
1951, 86-89. 

25. D. Katz, "Morale and motivation in industry" in Current Trends in Industrial 
Psychology. University of Pittsburgh Press, 1949, 145-71. 

26. D. Katz, "The Survey Research Center: An overview of the human relations 
program" in H. Guetzkow (Ed.), Groups, Leadership, and Men. Carnegie Press, 
1951, 68-85. 

27. D. Katz and H. Hyman, "Industrial morale and public opinion methods," Int. /. 
Opin. Att. Res., 1947, 1, 3, 14-30. 

28. D. Katz and R. L. Kahn, Some Recent Findings in Human Relations Research 
(Mimeographed Report). University of Michigan, Survey Research Center, 1952. 

29. D. Katz, N. Maccoby and N. C. Morse, Productivity, Supervision and Morale in 
an Office Situation. University of Michigan, Survey Research Center, 1950. 

30. D. Katz, N. Maccoby, G. Gurin and L. G. Floor, Productivity, Supervision and 
Morale Among Railroad Workers. University of Michigan, Survey Research Cen- 
ter, 1951. 


31. W. A. Kerr, "On the validity and reliability of the Job Satisfaction Tear Ballot," 
/. Appl. Psychol., 1948, 32, 275-81. 

32. W. A. Kcrr, Organizational Analysis of an Electronics Factory. Camden: Radio 
Corporation of America, 1943. 

33. W. A. Kerr, "Labor turnover and its correlates," /. Appl PsychoL, 1947, 31, 

34. W. A. Ken, "Where they lilce to work," /. Appl Psychol^ 1943, 27, 438-42. 

35. W. A. Ken, G. J. Koppelmeier and J. J. Sullivan, "Absenteeism, turnover and 
morale in a metal fabrication factory," Occup. Psychol, 1951, 25, 50-55. 

36. A. Kolstad, "Employee attitudes in a department store," /. Appl Psychol, 1938, 
22, 470-79. 

37. A. Kornhauser, "Psychological studies of employee attitudes," J. Consult. Psychol^ 

38. R. Lippitt, 'The morale of youth groups" in G. Watson (Ed.), Civilian Morale. 
Houghton Mifflin Co., 1942, 119-42. 

39. E. Mayo and G. F. F:' Lombard, Teamwork and Labor Turnover in the Aircraft 
Industry of Southern California (Business Research Studies No. 32). Harvard 
University, Graduate School of Business, Division of Research, 1944. 

39a. Q. McNemar, "Opinion-attitude methodology," Psychol. Butt., 1946, 43, 289- 

40. T. M. Newcomb, "News and morale: A miniature experiment" in G. Watson 
(Ed.), Civilian Morale. Houghton Mifflin Co., 1942, 175-85. 

41. D. L. Palmer, E. B. Purpus and L. Q. Stockford, "Why workers quit," Pen. /., 

42. K. Schooler, Methodology for Studying the Dimensions of Morale (Mimeographed 
Report) . University of Michigan, Survey Research Center, 1952. 

43. McG. Smith, "Mending our weakest links," Adv. Mgmt., 1942, 7, 77-83. 

44. R. G. Smith and R. J. Westen, Studies of Morale Methodology and Criteria 
(Research Bull., 51-29). San Antonio: USAF Air Training Command, Human 
Resources Research Center, 1951. 

45. S. Stouffer et al., The American Soldier: Combat and Its Aftermath (Vol. II. 
Studies in Social Psychology in World War II) . Princeton University Press, 1949. 

46. L. L. Thurstone, 'Theory of attitude measurement," Psychol. Rev. 1929, 36, 

47. L. L. Thurstone and E. J. Chave, The Measurement of Attitude. University of 
Chicago Press, 1929. 

48. Triple Audit of Industrial Relations (Industrial Relations Bulletin 11) . University 
of Minnesota, 1951. 

49. R. S. Uhrbrock, "Attitudes of 4,430 employees," /. Soc. Psychol., 1934, 5, 365-77. 

50. G. Watson, "Morale during unemployment" in G. Watson (Ed.), Civilian 
Morale. Houghton Mifflin Co., 1942, 273-348. 

51. G. Watson, "Five factors in morale," in G. Watson (Ed.), Civilian Morale. 
Houghton Mifflin Co., 1942, 30-47. 

52. R. J. Young, "Reduce excessive turnover costs through proper evaluation," Per- 
tonnel, 1950, 27, 75-79. 


The Pay Envelope 

FROM a review of problems in industrial relations, the con- 
clusion is reached by one observer that "the attitudes of too many com- 
panies toward human relations all too often seem to boil down to a 
non-expressed formula of 'give 'em a decent paycheck; that's all em- 
ployees are interested in'" (22). This fundamental generalization of 
traditional economics has been questioned by an increasing number of 
social scientists (4, 13, 18). The viewpoint of the latter finds striking 
expression in the conclusion, by Whitehead, that the basic psychological 
generalization of economics, which makes wages the major source of 
motivation and satisfaction in industry "has exhausted much of its use- 
fulness at the present time" and "has become a positive danger to our 
social and economic structure . . ." (28, p. 273). 


Conflicting opinions of the type cited above, in the face of wide-spread 
use of the wage incentive, underline the need for considering specifically 
what attitude surveys show concerning the potency of the pay envelope 
as an incentive to production and as a determinant of satisfaction and 
morale. In earlier chapters, reference has been made to the fact that 
both management executives and union leaders are inclined to the view 
that "higher wages" represents the prime, if not the sole, incentive to 
production. Thus, a survey conducted by the Opinion Research Corpora- 
tion in 1947, cited in Chapter 2, showed that 36, or 72 per cent of 50 
manufacturing plant executives held either the opinion that ( 1 ) "money 
alone is the answer" or that (2) "money is by far the chief thing" in 
stimulating workers to increased productivity (19). 

The possibility that executives tend to assign a degree of emphasis to 
compensation which large numbers of employees do not recognize is 
illustrated in an investigation, reported by S. J. Fosdick to the 1939 Con- 



vention of the National Retail Dry Goods Association (27). In this 
study, several hundred employers and 3,000 employees, representing a 
nation-wide sample, were asked to rank eight "morale'' factors. Credit 
for all work done, ranked first in importance by the employees, is seventh 
in the employers' list of rankings. Job security, rated as second in im- 
portance by employers, is identified as the least important factor by em- 
ployees. Management's tendency to consider pay as of prime (if not 
sole) importance is revealed in the selection of this as the most important 
factor by employers, as contrasted with the fact that employees rank this 
as third in importance. 

The fact that labor-union leaders as well as executives can and do err 
in their judgments a5 to what workers want is evidenced in the 1947 
study conducted by the National Industrial Conference Board (11), 
described in Chapter 13. Copies of a rating sheet for use in assessing the 
importance of 71 morale factors were forwarded to employees in six 
manufacturing plants and were also sent to the top policy-making ex- 
ecutives in the co-operating companies. Each of the latter group was 
asked to predict the rankings that would be assigned by the employees in 
his company. Similar rankings were obtained from 42 labor leaders to 
show what they believed represented the relative importance of the 71 
factors to their union members. ^ 

Comparison of rankings made by the three groups showed that ex- 
ecutives and labor leaders agreed in giving top ranking to Compensation 
base pay while Job security employment stabilization was given this 
position by the employees. In general, an overwhelming number of 
executives and labor leaders placed great emphasis on compensation 
while relatively few of the co-operating employees did so. In fact, ap- 
proximately 75 per cent of both executives and labor leaders included 
compensation (base pay) among their first 5 predictions while fewer 
than 30 per cent of employees included this among the five most im- 
portant factors. 


In a number of instances, pay has been assigned a relatively low posi- 
tion in employee ratings of factors associated with morale. Thus, as stated 
above, pay was rated third among eight morale factors listed in the 
study reported by the National Retail Dry Goods Association in 1939 
(27). Findings from a series of surveys conducted by Sears, Roebuck and 
Company over a 12-year period show that pay, as related to other jobs 
in the same unit, ranked eighth in place among the elements related to 
high morale. Rates of pay, as such, ranked in fourteenth place. Such 
findings are not interpreted by the investigators involved as meaning 


that rates of pay and proper job differentials are not important, since 
they loom up more importantly in employee thinking wherever local 
management is guilty of any serious shortcomings in these matters. How- 
ever, the position is taken that pay, as other material things, such as hours 
of work, are not enough in themselves; they are only the beginning. "If/' 
according to Worthy, "the only basis management can conceive for 
employee loyalty and cooperation is the pay envelope and the short 
work-week, there can never be enough money or short enough hours to do 
the job" (33, p. 65). 

The study of turnover and morale conducted by the Michigan Bell 
Telephone Company, described in Chapter 9, yields information con- 
cerning the attitudes of female operators and service representatives to- 
ward wages (30). An analysis of replies to the question "How (do) did 
you feel about the wages you receive (d)?" showed that the service repre- 
sentatives seemed less well pleased with their wages than the operators. 
Apart from this difference, however, according to Wickert, "there seems 
to have been some trend for the girls who left, whether they were opera- 
tors or service representatives, to feel better about their wages than those 
who stayed on!" x (30, p. 195). 

The view that money is not a powerful incentive in comparison with 
other features of the work situation is also taken as a result of an exten- 
sive survey of the attitudes of building operatives conducted in England 
and Scotland in 1946. Complaints about the inadequacy of wages, it was 
found, were widespread. Nearly half the men said that they could "just 
manage/' and approximately another one-third said that they "could not 
manage on their earnings" (6). Nevertheless, opposition to a system of 
payment by results was also widespread. Findings of interviews with 
400 workers, of a total of 2,300 employed on 14 sites, showed that nearly 
60 per cent of the men were, on the whole, opposed to this system of pay- 
ment and only 20 per cent expressed definite or conditional approval. 
According to Davis, the main arguments against a wage-incentive plan 
"were based on psychological grounds on the fear that this incentive 
would inhibit other strong and pleasurable motives for working, such 
as the pleasure of working for its own sake and the solidarity and good 
fellowship of the working group" (6, p. 59) . 


Contradictions Among Survey Findings: As is suggested in the discus- 
sion of factors to be considered in the interpretation of employee-attitude 
surveys, the significance of pay as an incentive has, at times, been con- 
cealed in the method used in the analysis of survey results (14). This, 

* Not italicized in the original. 


according to Wilkins, may account for the present tendency of some 
industrial psychologists and sociologists toward an undervaluation of the 
importance of pay as an incentive. "Perhaps/' he writes, "this is because 
they have relied too much on simple analysis and questioning. Whatever 
the cause, it would seem from any analysis of labor disputes and strikes, 
that rates of payment are very important in the workers' estimates, even 
when wage levels are above the subsistence margin" (31, p. 550) . 

The fact that pay is very much on the worker's mind can be made ap- 
parent from a consideration of survey findings discussed in Chapter 13. 
As shown in Figure 20, the number of unfavorable comments and the 
variety of unfavorable comments concerning pay were greater among 
employees of the Operating Branch, Western Electric Co., than for 
any other class of comments (17). When extra compensation plans 
(Factor 9) and profit-sharing plans (Factor 10) are considered along 
with Factor 7, i.e. compensation (base pay), it is found that approxi- 
mately 52 per cent of employees in the six companies surveyed by the 
National Industrial Conference Board place wages among the five most 
important morale factors (11). Approximately 41 per cent of employees 
participating in the General Motors Corporation's My Job Contest spon- 
taneously referred to wages as a source of satisfaction from the job and 
there was, in addition, approximately 22-per-cent mention of benefits 
from wages (9). 

Indications that level of pay exercises a significant influence on job 
satisfaction are found in a study, by Walker and Guest (25a), of 180 
workers employed on an automobile assembly line, representing a 
"stratified and substantial sample" of all productive workers in Plant X. 
Interviews, covering all phases of their life on "the line," were conducted 
in the workers' homes, starting in the summer of 1949. Through the use 
of a 52-item rating scale, comparative ratings were obtained of likes and 
dislikes pertaining to various elements of the total job situation. Economic 
factors, including pay and security, were given top rating in the like 
column of the rating scale. Of the 180 workers interviewed, 126 gave 
"good pay" as the primary reason for liking their job, and another 21 
rated "steady work" as first in importance. Furthermore, 81 rated "poor 
pay" or "work not steady" as the chief reason for disliking their previous 
jobs. The significance of these findings is increased by other data which 
showed that nearly 90 per cent of the men came to Plant X from jobs 
which were not machine-paced, a feature of the immediate job experience 
in Plant X which was most frequently checked in the dislike column of 
the rating scale. 

Evidence that workers believe that production can be stimulated by 
enlarging the "pay packet" is found in a survey (8) reported by the Opin- 


ion Research Corporation in 1951. Findings are based on a national 
sample of 1,245 employees of manufacturing companies, including manual 
and white-collar workers. As appears in Table 20, "more pay" is the an- 
swer most frequently given to the question "What should the company 
do to make the workers feel like turning out more worfe?" In addition, 
there is frequent reference to "incentive pay; bonuses." The data pre- 
sented in Table 20 are interpreted by the Opinion Research Corporation 
as indicating that "Higher pay tied to higher output is the most direct 
and obvious proof that higher productivity pays off.* But other types of 
rewards, even though less tangible, are clearly valued" (8, p. 52). 

The importance attached to monetary returns by the worker is also 
clearly demonstrated in the study by Rose (20), described in some detail 
in Chapter 18, involving a survey of a representative sampling (N = 392) 
of members of Local 688, International Brotherhood of Teamsters, Chauf- 
feurs, Warehousemen and Helpers of America, AFL (1), reported in 
1952. As shown in Table 33, 75.3 per cent of the workers interviewed 

TABLE 20. Workers 9 Opinion on Incentivation (after Public Opinion Index 
for Industry, Opinion Research Corporation, 1951 ) 

Answers to Question: "What should the company do to make the workers 
feel like turning out more work?" 

Chief categories Per cent Chief categories Per cent 

More pay 26 Longer hours 3 

Incentive pay; bonuses 21 Improved benefits 3 

Better employee relations 22 Encourage worker initiative 2 

Working conditions 14 Provide steady work 2 

Keep employees informed 7 Better recreation 2 

Better supervision 4 Rest periods 2 

Improved equipment 3 

mentioned "get specific economic benefits" (higher wages) as a purpose 
of the union. This is approximately two and one-half times the number 
of "mentions" of any other purpose. The results clearly typify the work- 
ers' concept of the economic role of the union and the importance of 
accomplishment in this direction as a contribution to union solidarity. 


The fact that workers want "more pay" at least in the same sense 
as they want more of whatever contributes to the gratification of wants 
appears in significant form in the findings of a survey made by the 
Office of Public Opinion Research (3). Each of 1,239 individuals, re- 
presenting a cross-sample of the national population, was asked to in- 

* Not italicized in the original. 



dicate the average weekly income group to which his immediate family 
belonged. Each was also asked, "About how much more money than 
that do you think your family would need to have the things that might 
make your family happier or more comfortable than it is now?" 

The relationships among actual income, income satisfaction, income 
aspiration are shown in Table 21. Of particular interest in this table are 
the indications that 

1. Over one-half of the population is dissatisfied with its present in- 

2. A large increase in income on the average, an 86 per cent increase 
over present income is desired by those who are dissatisfied and state 
how much more they want. 

3. The higher a person's income is, the more likely he is to be satis- 
fied with it and the smaller is the proportion of present income desired 
in way of an increase. 3 

4. For those persons who are dissatisfied, it is generally true that the 
more money a person has, the more money he wants. 

TABLE 21. Income Satisfaction and Income Aspiration in Relation to Present 
Income (after Centers and Cantril, 1946) 


Jo. specifying 
mount more 

* "Increase wanted? 





No a 


Dollars * 

Under $20 
$100 and over 








* The dollars-and-cents "increase wanted" is a translation based on the percentage 
indicated the mid-point of the income group concerned was used to construct this 
estimate. Ten dollars is used as the mid-point for the "under $20" group, while for 
the "$100 and over" group, no mid-point could be assigned, and thus $100 was used 
as the basis for calculation. 

The latter finding is of special significance in suggesting that an indi- 
vidual's present earnings provides him with a frame of reference for judg- 

1 An exception is noted in the case of the higher income brackets. According to the 
investigators, the data suggest that this occurs because at this level professional 
people are competing for social status with persons far above them in income. The 
relatively large increases desired are interpreted as representing a wish to strengthen 
an obviously high social status with a correspondingly high income status. Findings 
from studies by Wallace, Williams and Cantril (26) and by Smith (23) are cited in 
support of this view. 


ing his needs and for establishing his level of aspiration and goals. Pro- 
portionate to their present earnings, individuals in the lowest income 
brackets want a great deal more (increase wanted in percentages), but 
in comparison to the sums wanted by those in higher income brackets, 
increases wanted in dollars are smaller. 


The observation, in this study of adults, that expressed needs for 
amount of income is related to socio-economic station is supported by 
findings from a later study of high school students reported by Dittmer 
and Payne (7). Such studies perhaps merely confirm the widely held 
opinion that people at whatever level of income they may be want 
more money. A recent tendency has been to rationalize this want as 
representing not a direct motivational drive, but an urge toward the grati- 
fication of motives associated with status, recognition, and other ego- 
involved needs. This tendency appears not only in the consideration 
of direct wages as an incentive, but in current discussions of other eco- 
nomic benefits. For example, Sears, Roebuck and Company has a broad 
program of employee benefits, including group insurance, illness allow- 
ances, severance pay, and other benefits voluntarily provided by com- 
pany policy. This program, together with all the items customarily in- 
cluded under the heading of "nonwage payments" involved an expendi- 
ture in excess of 84 million dollars in 1948. Survey findings show that 
this framework of employee benefits has had a definite influence on em- 
ployee morale. However, according to Worthy, "Important as the eco- 
nomic aspects of the benefits may be, their chief significance as regards 
employee morale is the fact that they are tangible evidence to the or- 
ganization of management's concern for the welfare of its employees . . . 
Our studies show clearly that employees respond primarily to the evidence 
of management's concern and only secondarily to the economic values 
as such" (33, p. 66). 

Included among benefits available to Sears, Roebuck and Company 
employees is a profit-sharing plan (10) which, in continuous operation 
since 1916, is open to all employees with a year or more of service. Each 
member of the plan deposits 5 per cent of his earnings (up to a maximum 
deposit of $250 per year) in the profit-sharing fund. At the end of the 
year the company contributes to the fund between 5 and 9 per cent (de- 
termined according to a fixed scale) of its net profits before taxes and 
dividends, representing in 1948 a contribution of $22,817,079, credited 
to the accounts of employees on a dual basis of length of service and 
amount of individual deposit. In 1950, it was reported that the profit- 
sharing fund held in trust for its members represented approximately 19 


per cent of the company's outstanding capital stock, making Sears em- 
ployees as a group by far the largest stockholders in the company. 

According to Worthy, what has been said above concerning employee 
benefits applies particularly to the company's profit-sharing plan. "Over 
the years the profit sharing plan has gradually assumed a role which 
transcends its function as tangible evidence of management's concern 
for the welfare of its employees. In a sense, profit sharing has become a 
'unifying principle' that serves as a symbol around which the entire or- 
ganization revolves . . . Here again/' adds Worthy, "the influence on 
employee attitudes is primarily symbolic rather than direct. The actual 
holdings of individual employees are not nearly so important as the fact 
that the huge holdings of the profit sharing fund symbolize to them that 
they and the company are one, that they are in effect 'working for them- 
selves' . . ."(33, p. 67). 

In this respect, the view expressed by Worthy coincides with that taken 
by the president of the Lincoln Electric Co., who reports that the chief 
effect of its incentive-management plan, which includes profit sharing, 
has been to change "workers from people who are working at a job at so 
much an hour into people who feel their success is tied definitely, com- 
pletely and proportionately into the success of the company itself" (25). 
The fundamental effect of this- plan, it is said, "is the development 
of the individual through his desire to rise in his usefulness in industry 
and through his desire to work together with his co-workers from top to 
bottom to produce better products more efficiently" * (15, p. 11). 

In the light both of studies reported in this volume and of everyday 
experiences, there can be no disagreement with the position, early stated 
by Whiting Williams, that "it is impossible ... to judge the effect of 
either wages or other conditions of work apart from the relationships 
the work permits with other persons" (32, p. 62) both in the plant and 
the community at large. It would be shortsighted to disregard the sig- 
nificance of earnings as a symbol of status. However, recognition of what 
Williams has called the "social handles of the pay-cup" should not lead 
to an underestimation of the potential of the financial reward in itself 
as an incentive to production. 


As shown in earlier pages of this volume, particularly in Chapter 2, 
day-to-day experience in industry, as well as investigations bearing upon 
wage-incentive plans, show that the pay incentive can serve to increase 
earnings, raise production, and reduce costs in spite of adverse employee 

4 No italics in the original. 


ittitudes and even of widespread opposition to the use of such plans 
( 16, p. 277) . While good results can be attributed to wage-incentive plans, 
there is also reason for the opinion that much remains to be done to in- 
sure the optimal utilization of the pay envelope as an incentive to pro- 
duction (5) and in building employee morale. The basic problem of 
what constitutes a "fair day's pay" (24) is still far from complete solu- 
tion, in spite of the best efforts of the economists on the one hand and 
of those concerned with job evaluation and wage structures for indi- 
vidual plants on the other. While there has been considerable improve- 
ment in recent years in the construction and administration of wage in- 
centive plans, technical problems in this area are far from solved. 

More important still, as Whyte has pointed out, is the need for recog- 
nizing that "financial incentives are both a technical engineering and a 
human relations problem." Thus, according to Whyte, 

"Many people are arguing the question today: Which is more important 
to workers, economic incentives or human relations? In that form the 
question is meaningless and unanswerable. Men are interested in money. 
They are also concerned about their relations with other men. Offer 
them a financial reward for behavior that damages their relations with 
other men, and you can hardly expect them to respond with enthusiasm" 
(29, p. 73). 

The issue, Whyte goes on to say, is nevertheless not one of choosing 
between economic incentives and human relations. "The problem is to 
fit economic incentives and human relations together, to integrate them" 
(29, p. 73). Methods for achieving such integration are suggested, in 
part, in studies described in Chapter 9, which show that employee par- 
ticipation in setting goals heightens the effectiveness of a wage-incentive 
system. In general, it has become evident that the effective use of the 
wage incentive requires a thorough understanding of the attitudes both 
of small informal work groups and of formal employee organizations 
(1, 12). Much depends also upon the spirit with which wage incentives 
are approached, the "climate" within which wage incentive plans are 
installed (16). Harmonious human relations reflected, in part, through 
the formal employee and labor relations programs of the plant con- 
stitutes a basic ingredient of a "climate" suitable for increasing the "pull- 
ing power of the money reward" (29). However, the necessity for meet- 
ing such requirements does not call for the belittlement of the pay 
envelope, but rather for systematic action designed to enhance its value 
as an incentive to production and as a fruitful medium for the satisfaction 
of workers' wants and needs. 



1. S. Barkin, 'Trade-union attitudes and their effect upon productivity" in L. R. 
Tripp (Ed.), Industrial Productivity. Madison, Wis.: Industrial Relations Research 
Assn., 1951, 110-29. 

2. W. D. Brown, "Incentives within the factory/' Occup. Psycho/., 1945, 19, 82-92. 

3. R. Centers and H. Cantril, "Income satisfaction and income aspiration," /. Ab- 
norm. Soc. Psycho/., 1946, 4, 64-69. 

4. A. Curie, "The sociological background to incentives," Occup. Psycho/., 1949, 23, 

5. M. Dalton, "Economic incentives and human relations" in L. R. Tripp (Ed.), 
Industrial Productivity. Madison, Wis.: Industrial Relations Research Assn., 
1951, 130-45. 

6. N. M. Davis, "Attitudes to work among building operatives," Occup. Psycho/., 
1948, 22, 56-62. 

7. R. M. Dittmer and S. L. Payne, "Living Research," Publ Opin. Quart., 1948, 12, 

8. Employee Cooperation on Productivity. Princeton: Opinion Research Corpora- 
tion, 1951. 

9. C. E. Evans and La V. N. Laseau, My Job Contest (Personnel Psychology Mono- 
graph No. 1). Washington: Personnel Psychology, Inc., 1950. 

10. "Every man a capitalist," Time, Feb. 2, 1952, 90. 

11. Factors Affecting Employee Morale (Studies in Personnel Policy No. 85). New 
York: National Industrial Conference Board, 1947. 

12. W. A. Gomberg, A Trade Union Analysis of Time Study. Chicago: Science Re- 
search Associates, 1948. 

13. S. P. Hayes, Jr., "Some psychological problems of economics," Psycho/. Bull, 
1950, 47, 289-330. 

14. A. Kornhauser, "Psychological studies of employee attitudes," /. Consult. Psycho/., 
1944, 8, 127-43. 

15. J. F. Lincoln, Incentive Management. Cleveland: Lincoln Electric Co., 1951. 

16. H. B. Maynard, "Changing philosophies on wage incentives," Mech. Eng., 1952, 
74, 277-79, 314 ff. 

17. E. Mayo, The Human Problems of an Industrial Civilization. Harvard Business 
School, Division of Research, 1946. (First printed by The Macmillan Co., 1933.) 

18. R. Marriott, "Socio-psychological factors in productivity," Occup. Psycho/., 1951, 

19. Productivity and the Worker. Princeton: Opinion Research Corporation, 1947. 

20. A. M. Rose, Union Solidarity. University of Minnesota Press, 1952. 

21. M. Sherif and H. Cantril, The Psychology of Ego-Involvements. John Wiley 
and Sons, Inc., 1947. 

22. H. G. Simpson, " 'Humanized' industry a management 'must/ " Forbes, ( Septem- 
ber 1) 1948,62,155. 

23. R. B. Smith, "The development of an inventory for the measurement of inferior- 
ity feelings at the high school level," Archiv. Psycho/., No. 144, 1932. 

24. M. W. Reder, "The general level of money wages," in Proceedings of Third An- 
nual Meeting, Industrial Relations Research Assn., Chicago, 111., 1950, 186-202. 


25. Wage Incentive Plans (Studies in Personnel Policy No. 68). New York: National 
Industrial Conference Board, 1948. 

25a. C. R. Walker and R. H. Guest, The Man on the Assembly Line. Harvard Uni- 
versity Press, 1952. 

26. J. M. Wallace, Jr., F. W. Williams and H. Cantril, "Identification of occupa- 
tional groups with economic and social class/' /. Abnorm. Soc. PsychoL, 1944, 39, 

27. G. Watson, "Work satisfaction" in G. W. Hartmann and T. Newcomb (Eds.), 
Industrial Conflict. The Cordon Co., 1939, 114-24. 

28. T. N. Whitehead, "Social motives in economic activities/' Occup. Psycho!., 1938, 
12, 271-90. 

29. W. F. Whyte, "Economic incentives and human relations/' Harvard Bus. Rev., 
1952, 31, 73-80. 

30. F. R. Wickert, "Turnover and feelings of ego-involvement/' Pers. Psychol., 1951, 
4, 185-97. 

31. L. T. Wilkins, "Incentives and the young male worker/' Int. J. Opin. Att. Res., 

32. W. Williams, Mainsprings of Men. New York: Charles Scribner's Sons, 1925. 

33. }. C. Worthy, "Factors influencing employee morale," Harvard Bus. Rev., 1950, 
28, 1, 61-73. 


The Need (or Security 

THE HISTORY of the Industrial Revolution is replete with 
instances of destructive attacks by workers on machines which they felt 
would deprive them of their jobs (56). Today, v/hile seldom destroy- 
ing the machines, workers take other steps, ranging from "slowdowns" 
to actual sabotage to prevent or hinder the introduction of technologi- 
cal and other changes in industry. According to Barkin, a spokesman for 
organized labor, the current mood of labor, as contrasted with the pre- 
vailing climate of the 1930's, is~one which is comparatively favorable to 
change (6, 8). Nevertheless, in a review of the outlooks of both the 
owners and employees toward technological change, Barkin adds that 
no group in American society is free of opposition to changes which 
will affect its members adversely and, to protect their modes of life, all 
groups can be expected to devise special defenses and show resistance 
to innovation (6). 

In recent years there has been an increasing tendency on the part of 
social scientists to explain such attitudes and conduct largely on the 
ground that resistance to change is deeply ingrained in the human race 
(43). While ingrained habits undoubtedly contribute to resistance to 
change, there seems no question that such resistance on the part of work- 
ers also has its roots in basic anxieties aroused by conditions of the work 
situation. One prime cause of such anxiety is the fear of losing the job in 
the face of a deep-seated need for security in terms both of the immediate 
situation and of outlooks for the future (44). 


Many attitude surveys have shown that need for security is widespread 
and deep-seated. As indicated in Table 17, approximately 31 per cent of 
employees in the six plants surveyed by the National Industrial Con- 
ference Board in 1947 ranked "Jb security-employment stabilization'' as 


Erst in importance among 71 morale factors. This percentage is three 
times as high as that applying to "Compensation-base pay" (8.7 per cent) 
which was second in frequency of choice as the most important factor 
(23). Reference to Table 18 shows that about 50 per cent of the em- 
ployees involved selected factors related to Security as being among the 
five most important factors. There is also considerable evidence from 
other attitude surveys to support the view that economic security, as re- 
presented in steady worfe and regular compensation, is a basic need of 
the worker. 

Attitudes of High School Children: That this need appears early in the 
life of the individual is suggested in a study conducted in 1949 by the 
Social Research Service, Michigan State College, involving 6,789 tenth- 
and twelfth-grade students from 56 Michigan public and private high 
schools. Data presented in Figure 26 lead to the conclusion that "high 
school students reflect the experience of many American workers" in their 
unwillingness "to accept an insecure position for the sake of high income" 
(61, p. 11). 

Attitudes of Applicants for Employment: Confirmatory evidence with 
respect to attitudes of young people toward steady work and regular com- 
pensation is found in an investigation by Jurgensen (31) who, in 1945-46, 
asked 1,189 male applicants for employment by the Minneapolis Gas 
Company to rank ten factors in order of importance with respect to 
"Job Preferences (What makes a job good or bad?)! 9 Items which were 
rated included (1) Advancement; (2) Benefits; (3) Company; (4) Co- 
Workers; (5) Hours; (6) Pay; (7) Security; (8) Supervisor; (9) Type of 
Work, and (10) Working Conditions. 

Security, with a mean rank of 3.1, proved to be the most preferred job 
characteristic, followed closely by advancement (mean rank 3.4) and 
type of work (mean rank 3.7). Pay is sixth in the list (mean rank 6.3) 
with working conditions and benefits (mean rank 7.3) matched for last 
place. A later analysis of replies by nearly 3,345 male applicants (32) 
yielded similar findings, with security, opportunity for advancement, and 
type of work again heading the list of preferences, in the order named. 

A Nation-Wide Sample of Workers: A survey conducted in 1947 by 
the Opinion Research Corporation, involving a nation-wide survey of 
employed workers, shows, as indicated in Figure 27, that the latter agree 
with high-school students and applicants for employment in rating se- 
curity high as a source of satisfaction on the job. The results of the study 
lead to the suggestion that in so far as "job satisfaction is partly a func- 
tion of interpretation or how the job is sold to employees, it follows that 
management can stimulate job satisfaction by showing what it is doing 
to make jobs more secure" (16) . 




Here are three different kinds of jobs. It I had 
my choice, / would pick a job which pays: 

Quite a low 
income but 
which 1 was 
sure of keeping 

A good income 
but which 1 
hove a 50-50 
chance of losing 

An extremely 
high income 
if 1 con moke 
the grade but 
in which 1 
lose almost 
everything if 1 
don't make it 

No response 



| 43% 



|f j 48% 

f (44% 


j f 8% 


j 3% 


Figure 26. Attitudes of High-School Students toward Security and Income * 

A Study in a New England Manufacturing Town: An analysis of 
factors affecting choice of jobs was included in a survey of two samples 
of manual workers, totalling about 800 individuals, in a medium-sized 
New England manufacturing town (47). When asked to express a pre- 
ference between a "wage increase" or a "guarantee of steady employ 
ment," 73 per cent chose the latter. 

Department-Store Employees: The Jurgensen Job Preference Blank 
has been used by Hardin, Reif and Heneman in an investigation, in 1949, 

* By permission, from Youth and the World of Work, Michigan State College, 
















Figure 27. What Do Men Value in Their Jobs? * 

of the job preferences of employees in two midwestern department stores 
(28). As in the case of applicants for employment by the public-utility 
company, security (with a mean rank of 3.3) proved to be the most pre- 
ferred job characteristic, again followed by advancement (mean rank 3.8) 
in rankings by male employees (N = 63). Security was also given the 
first position by women (N = 139) with type of work (mean rank 3.8) 
following in second place. Pay was given fourth place by men and third 
place by women. An interesting additional outcome of the study was the 
finding that test-retest reliability for mean rankings was high (0.98), al- 
though individual preference factors showed more inconsistency than did 
group factors. 

Building-Trade Workers in Great Britain: Further evidence concerning 
the importance attached to security is found in the British survey of 
building-trade workers, referred to in Chapter 15. The survey was con- 
ducted in a period (1946) when fear of unemployment was less acute 
than in earlier years. Nevertheless, according to the investigator, "vivid 
memories of the past and uneasy scepticism about the future made the 
assurance of greater security and a guaranteed week seem a matter of 
urgent importance to the operatives" (18, p. 57). The theme of security, 
it is reported, recurred throughout the interviews, and 16 per cent of 
those interviewed reported that they felt "definitely insecure/' while an- 
other 24.3 per cent were "doubtful." 

Union Members: Of union members surveyed by Rose (49) in the 
study described in Chapter 18, 31.1 per cent mentioned "get job security" 

* By permission, from Public Opinion Index for Industry, Opinion Research Cor- 
poration, 1947. 



(including seniority) as a purpose of the union. In order of frequency of 
mention, this item stands second in a list of eight stated purposes, and is 
accorded twice as many mentions as the next most frequent item on 
the list, viz., "gain rights" (e.g., welfare, free speech, fair deal) . This may 
be taken as further evidence that workers feel a deep-seated need for se- 


The Morale of Unemployed Persons: A number of studies have shown 
that insecurity, characteristic of unemployment, severely affects the atti- 
tudes of workers (55). Evidence of this kind is found in a study con- 
ducted by Hall (26) during the severe economic depression of the mid- 
1930's in the United States, Subjects of the study were employed and un- 
employed engineers. A morale score was obtained for each of these by 
scoring five items on a questionnaire of the Likert type. Corrected co- 
efficients for split-half reliability were found to vary from .69 to .84 and 

Unemployed Meq 

A* Men la this group are unemployed 
but are in no Immediate need of, 
financial help* (N=91) 

C. Men who, like those in group D, 
were in desperate straits, but 
had been given work relief l (N=114l 

Average, all unemployed man 


B. These men are receiving help from 
family or friends. (H*105) 

D* Men vho, like those in group C, 
vere in desperate straits but vho 
had not been given work relief * 





Employed Hen 

! Men in this group feel that 

their Jobs are secure. (N=118) 

** Average, all employed men 
^ (N*500) 

;2 These men sometimes feel anxiety 
that they may be laid off.(N=108) 

5. These men anticipate being laid 
off almost any time* (N=74) 

Low Morale 

Figure 28, Average Occupational Morale Scores of Groups at Various Levels 
of Security * 

* By permission, from O. M. Hall, "Attitudes and unemployment/' Arcfitv. Psyc/to/., 
No. 165, 1934. 


retest after 30 days yielded a correlation of .63 with original scores. As 
shown in Figure 28, morale deterioration increases with increasing anxiety 
about having a job and with increasing economic insecurity. Studies by 
Bakke (3, 4), Rundquist and Sletto (50), Zawadski and Lazarsfeld (62), 
and by others have also revealed changes in attitudes, morale and person- 
ality with continued unemployment which, when carried into the career 
of the individual subsequent to unemployment, can be detrimental from 
the viewpoints of production and adjustment at work, and in terms of 
the wider problem of industrial strife. 

The Effects of Anticipated Loss of the Job: A more recent investiga- 
tion by Grove and Kerr has shown further that a feeling of insecurity, 
resulting from anticipated loss of employment can produce an "inverse 
halo" effect which apparently causes employees "to express discontent 
with their actually superior pay and working conditions as well as 'lower 
than normal expectancy' attitudes toward even their work associates and 
immediate supervisors" (25). 

Evidence for this conclusion is found in a study in which the Tear 
Ballot for Industry, General Opinion (35), providing measures on ten 
employee morale variables, was administered to 18 female office workers 
in an office which had entered receivership six months previously. Dur- 
ing the six-month period surviving employees saw fellow workers being 
laid off each week and economy measures adopted. At the same time, 
pay, working conditions, etc., in the plant remained superior to those gen- 
erally found in the area. The same questionnaire was applied to a con- 
trol group of 58 female office workers in a neighboring plant with a finan- 
cially sound outlook, where no significant layoffs had taken place, and 
where both salaries and working conditions were average. 

In Table 22 are shown the percentage of workers in each of the two 
groups reporting the two "most favorable" of the five possible levels of 
satisfaction on each of the ten job morale items. All percentage differences 
between the "insecure" and "normal" groups are statistically significant 
at the 5-per-cent level or better. The largest difference in the proportion 
giving favorable responses is found in the case of "job security," but it 
is evident that markedly lower morale in the insecure group is reflected 
in all of the items. Analysis of findings leads to the conclusion that, be- 
yond reasonable doubt, some kind of reverse halo effect is operating and 
that the generator of this is job insecurity. According to Grove and Kerr, 
44 Apparently, in these 18 cases, the job insecurity experience has, in time, 
so permeated employee feeling tone toward the entire job situation that 
the fear-shock has seeped like a toxic inhibitor into all job-related atti- 
tudeswith, perhaps, the original cause of the fear being manipulated 
stiffly in thought merely as a logic-tight symbol" (25, p. 169) . 


TABLE 22. Per Cent Favorable Response of Female Office Workers in a 

Bankrupt Firm Compared with Responses of a "Normal" Group of 

Female Office Workers (after Grove and Kerr) 

Normal Insecure 

Job security 88 17 

Welfare interest 78 22 

Ability of superior 50 28 

Working conditions 47 11 

Associates 98 72 

Wages 21 12 

Open-door management 67 23 

Management good intent 81 28 

Management good sense 71 17 

Personal happiness 66 6 


The studies cited above, conducted during the 1940's and later, yield 
findings in agreement with those resulting from investigations made in 
the 1930's by Chant in the United States, and by Wyatt and Langdon 
in England, in which steady work was rated high and considerably above 
high pay among 12 factors by 100 department store workers, 150 miscel- 
laneous workers (13) and 325 factory workers (60). However, other 
studies conducted during the same periods yield results which appear to 
contradict those cited above. For example as can be seen from Figure 20, 
the frequency of comments on steady work is extremely low in the case 
of employees of the Operating Branch, Western Electric Company, 
interviewed in 1929. Job security ranks lowest in importance among eight 
morale factors rated by 3,000 retail-store employees surveyed by Fosdick 
in 1939, although their employers rank job security as second in impor- 
tance (54). Information supplied to the National Industrial Conference 
Board (1947) by Stech Associates, based on the opinions of several thou- 
sand employees in 13 organizations, gives "job security compared with 
elsewhere' 9 a rank of 20 in a list of 32 morale factors (23) . 

In 1945, Centers (12) asked 1,071 men, constituting a representative 
cross-section of the adult white male population (including the complete 
occupational stratification), "Are you satisfied or dissatisfied with your 
present job?" Of this group, 85 per cent were classified as satisfied and 
15 per cent as dissatisfied on the basis of their replies. The former were 
asked "What is it that you like about the job?" "Security, steady work" 
was mentioned by only 6.6 per cent of the respondents, and this factor 
is 6th in a list of replies which was topped by such items as "freedom, in- 
dependence" (25.5 per cent); "interesting and varying activities" (20.6 
per cent); "pay" (8.9 per cent) . On the other hand, the responses of dis- 


satisfied individuals to the question ''What is it that you don't like about 
the job?" placed the reply "uncertainty, insecurity" second in a list (13.3 
per cent) preceded, however, by the reply "low pay, poor profits" given 
by 21.5 per cent of those interviewed. 

Of additional interest, in considering the role of the steady job and 
regular pay in moulding employee attitudes and in gratifying motives 
are findings from the General Motor Corporation's My Job Contest (22). 
If all themes mentioned which, in the opinion of this author, can logically 
be subsumed under a heading of security are considered, it is found that 
the per-cent mention of such themes is 67.3 (Table 18). However, for 
purposes of M/C coding only comments which mentioned security as 
such, or attributed a feeling of security as a result of some specific 
factor, were classified under the heading of "security." Examples of such 
classifications include the statements "I can always count on that check 
when I need it" and "the life insurance policy partly paid by the company 
gives me a feeling of security" Applying this standard, the investigators 
found that only 21.5 per cent of M/C participants mentioned "security" 
as a theme. Comparison of findings in the various divisions of the com- 
pany suggested further that "security" was not particularly sensitive as 
a theme for differentiating among divisions with respect to their per- 
formance in the area of employee relations. The conclusion is drawn that 
"identification with General Motors in itself gave some a feeling of 


It is apparent, from the above illustration, that discrepancies with re- 
spect to the importance attached to security may reflect difference in the 
"frames of reference" employed in the interpretation of results (34, 37, 
52). Other factors, of the type discussed in Chapter 12, may be involved. 
For example, some of the surveys described above were made in the late 
1930's, during a period of depression; others were conducted in the mid- 
1940's during a period of prosperity in the United States. It may be 
postulated that differences in general economic conditions between such 
periods may help account for the discrepancies in findings that "steady 
work" would move upward in an order of preference ranking during a 
period of depression (20). However, an examination of the data does 
not appear to support this assumption. 

The Influence of Socio-Economic Status: Differences among the popu- 
lations covered by the separate investigations may account for the dif- 
ferences in findings. Data presented by a number of investigators show, 
in fact, that various segments of both the adult working population and 
of high-school students respond differently to questions pertaining to 


the need or desire for economic security. For example, in a study by Korn- 
hauser (36), made in 1936, 350 male subjects in Chicago were classified 
into four income groups viz., wealthy, upper, middle, and lower. Results 
showed that security and independence were most strongly desired by 
the lower and middle income groups. In the survey conducted by the 
Social Research Service, Michigan State College in 1949, the percentage 
emphasizing security, as contrasted with size of income on a job, is higher 
among students whose fathers are unskilled or semiskilled workers than 
among those who are the children of parents engaged in professional 
or proprietary occupations (61). Similarly, Centers (12) found that 14.5 
per cent of unskilled workers and 9.2 per cent of semiskilled workers inter- 
viewed by the staff of the Office of Public Opinion Research in 1945 
replied ''Security, steady work" to the query "What is that you like about 
your job?" as contrasted with 1.7 per cent of professional workers and 5.9 
per cent of skilled workers. 

In one phase of the study by Centers, each of the individuals inter- 
viewed was handed a card containing the statements shown below and 
asked "If you had a choice of one of these kinds of jobs, "which -would you 
choose? Just call out the letter" 

A. A job 'where you could be a leader. 

B. A very interesting job. 

C. A job where you would be looked upon very highly by your fellow 

D. A job where you could be boss. 

E. A job which you were absolutely sure of keeping. 

F. A job where you could express your feelings, ideas, talent, or skill. 

G. A very highly paid job. 

H. A job where you could make a name for yourself or become famous. 

When the first two choices were combined to give an index of value, 
the percentage of all manual workers choosing security proved to be 37.7 
per cent; * of all unskilled workers 47.2 per cent; of all business, profes- 
sional and white collar workers, 16.6 per cent. The findings, as discussed 
further on pages 381-382, showed fairly consistent tendencies for the 
desire for security to increase as first choice as lower and lower occupa- 
tional levels were scrutinized. According to Centers, "people in the lower 
strata of society appear dominated by the simple problems of existence 
as expressed in the desire for security, that is, for a job they could be ab- 
solutely sure of keeping, but the higher occupational groups seem rela- 
tively free of such a basic need . . ." ( 12, p. 214) . 

The Influence of Intelligence: The possibility that differences in Intel- 

1 Maximum possible value equals 200 per cent, since each person included chose 
two situations. 


ligence, as well as in occupational levels, affect attitudes towards eco- 
nomic security and the role of steady work as an incentive, is suggested 
in studies by Wilkins (57, 58) involving 300 men, between the ages of 
18 and 19, interviewed at Army Reception Centres throughout Great 
Britain. Correlations between ranks assigned to eight factors, designated 
as "incentives'* by the investigators, were subjected to factor analysis 
by the centroid method. When thus analyzed, the 8 factors divide into 
two equal groups as shown in Table 23. 

TABLE 23. Classification of Incentives by Factorial Analyses (after Wilkins) 

Group A Croup B 

Prospects Workmates 

Security Hours 

Variety Pay 

Efficient organization Leave 

According to Wilkins, the following arbitrary labels may be applied 
to these divisions 

A Factors of long-term appeal, associated with the job itself. 
B Factors incidental to the job itself, and of shorter-term appeal. 

Comparisons of groups classified by intelligence showed further that 
higher intelligence groups are more interested in the items classified under 
"A" above; they want to know about the "job itself' and not the "com- 
pensations for doing it." According to Wilkins, "in general, the higher 
intelligence groups seek long-term incentives closely associated with the 
job itself, and the lower intelligence groups seek shorter-term incentives 
and tend to look more to compensations for doing the job, or to social 
incentives" (58, p. 560). 

Providing Security for Employees 

Consideration of the varied and sometimes contradictory findings does 
not justify the conclusion that economic or job security is the prime 
need of all workers or that "job security is the heart of the labor relations 
problem" (23, p. 19). The findings also fail to support the viewpoint, 
implied in some treatments of the problem, that the need for job se- 
curity is an artifact growing out of seniority regulations, featherbedding, 
labor-relations laws and political directives (16). The outcomes of em- 
ployee attitude surveys support the conclusion that economic security, 
as represented in steady work and regular compensation, is a basic need 
of the worker which must be considered in the formulation of personnel 
policies and practices in industry and business. Findings further indicate 


quite clearly that the need for economic security is particularly strong 
among employees in lower level occupations and suggest that the failure 
to satisfy this need can become a major source of dissatisfaction at these 


It would appear that, in the case particularly of such groups, steps taken 
by industry to provide reasonably steady work and regular income can pro- 
duce significant returns in the way of improved morale and motivation. 
It may even be that, as suggested in the report of the 1947 National In- 
dustrial Conference Board survey, "nothing that a company can do fur- 
nishes stronger and more tangible evidence of genuine concern for the 
welfare of its employees than an exhaustive attempt to provide a steady 
income for as many of its workers as possible" (23, p. 19). 

The Viewpoints of Union Leaders: The central importance of job se- 
curity is being increasingly stressed by union leaders (29, 49). Thus, ac- 
cording to Barbash, "If there is one overriding consideration which moti- 
vates the unions' attitudes with respect to the terms of the collective 
agreement it is fear fear of there not being enough jobs to go around, 
fear of unjust discharge, and fear of unequal treatment in short, inse- 
curity . . ." (9, p. 99). According to Barkin, a spokesman for organized 
labor, one function of "deliberative procedure in collective bargaining" 
is to help workers accommodate themselves to change, as one means of 
preventing the worker's displacement as a result of technological and 
other changes (6). Barkin also insists (5, 7) on full presentation of data 
by management and a complete review of proposed changes by the union 
so that the union can obtain "maximum protection for their members 
in the light of their own concepts of negotiable limits . . ." Among the 
objectives of such review and negotiation, it is said, will be the retention 
of the older skills and job patterns in so far as possible; guarantees against 
displacement of employees; reduction of new hirings until regular em- 
ployment is assured displaced employees; preferences for displaced em- 
ployees to all new jobs, etc. (6, p. 124). 

Action taken in the Fall of 1951 by CIO leaders in preparing for nego- 
tiations with "Big Steel" and "Little Steel" shows clearly the growing 
importance attached by union leaders to the satisfaction of the need for 
economic security. It seems likely that demands for the "year-around pay 
system" is to become a major item in union policy (30). Thus, in 1952, 
Walter Reuther unequivocally stated: "Our next basic demand will be 
for guaranteed annual wages. You could have the highest hourly wages 
in the world, but if you are not working they are of little value. People 
who get paid by the year have what we consider a very convenient ar- 


rangement They eat by the year and get paid by the year; but a worker 
who gets paid by the hour and eats by the year has a problem. We shall 
seek an arrangement for a guaranteed annual wage, a practical tool by 
which we may achieve a balance between our ability to create wealth and 
our ability to consume the wealth we know how to create" (48, p. 93 ).* 

Management Viewpoints and Responsibilities: According to the Na- 
tional Industrial Conference Board: "Trends in recent years on the politi- 
cal front indicate that the challenge to provide security can neither be 
evaded nor refused. In fact, there is increasing acceptance of the tenet 
that if private business fails to meet the challenge it will yield its leader- 
ship to some group that can. The question no longer appears to be, 'Shall 
the worker's sense of security be increased?' but rather, 'From what source 
will it emanate?' " (23, p. 19). 

Increases during recent years in the number of stabilized employment 
plans, in the way of guaranteed annual wages or in other forms, indicate 
growing recognition of the problem of economic security on the part 
of industry. 8 Progressive management has, in fact, taken the position 
that this is one of the prime responsibilities. Thus, according to the presi- 
dent of the Lincoln Manufacturing Co., "It is management's duty to 
make the worker secure in his job . . . The fact that the usual manage- 
ment will fire the worker when he runs out of work has had more to do 
with production limitation than any other circumstance. No man will 
willingly work to throw himself out of his job, nor should he" ( 39, p. 170) . 

There are many reasons apart from fear of increased union influence 
and government interference in business for management interest and 
action in dealing with the problem of security. As shown on earlier pages, 
a feeling of insecurity results in job dissatisfaction and lowered employee 
morale. Job insecurity contributes to job inefficiency, since the desire 
for continued employment helps to account for organized and unor- 
ganized employee resistance to increased productivity and technologi- 
cal progress (19, 40, 41, 42). Furthermore, according to the National In- 
dustrial Conference Board, employees who lack security will tend to 
spend working time in "jockeying for position" for the purpose of gain- 
ing a strategic position favorable to security. Rumors, bickering, conflicts 
with fellow workers, an increased number of grievances are associated 

2 At a UAW convention held in March, 1953, a huge sign stretched across the 
meeting hall read NEXT STEP FORWARD GUARANTEED ANNUAL WAGE. Walter Reuther 
is quoted as announcing: "We say to the employers: 4 We are not going to sign new 
[contracts] until you put into them guaranteed annual wages for the workers in our 
basic industries' '' (Time, April 6, 1953, 98). 

8 For references on stabilized employment and guaranteed annual wages see Feld- 
man (24), Chernick (14), Chernick and Hellickson (15), Pigou (46), Kaplan (33), 
Thompson (53), Snider (51), and others (1, 2, 21, 21a). 


with the struggle for security. "Where/' according to the National In- 
dustrial Conference Board, "the chief executive has failed to recognize 
insecurity as the basic cause, his attempts to achieve the degree of effi- 
ciency of which the organization is capable has aggravated the situation 
rather than improved it." On the other hand, "where security is achieved 
and the employee is stimulated by effective leadership, efficiency increases 
through the willing application of the worker to the job" (23, p. 20). 


Emphasis in the preceding discussion has been primarily upon job or 
immediate security. Concern for long-range security is revealed in cur- 
rent union demands fot pensions and for benefits for employees during 
temporary unemployment because of illness or for other reasons. Such 
demands particularly in the case of pensions appear to be consistent 
with the expressed desires of workers (Table 18). 

Evidence of this is found in a survey reported by Modern Industry in 
1950. This showed that there is widespread recognition among wage 
earners that government Social Security alone will not provide enough 
for self-support during old age. Most workers also indicated that they 
had been unable to "tuck away" , savings for retirement purposes from 
their earnings. Those interviewed were asked "As you see it, whose re- 
sponsibility is it for making up the difference between Social Security and 
what you will need in addition?" According to 25 per cent of those inter- 
viewed, this is the sole responsibility of the company for which they 
work. An effort was made in the survey to obtain workers' evaluation of 
pensions in companies with other benefits by asking "I/ you had to choose 
between a pension at 65 or any one of the following benefits which would 
you rather have?" As can be seen from Table 24, when workers are asked 
to choose between pensions at 65 and a number of other benefits, "pen- 
sions have a clear lead over any one other benefit/' From these and other 
findings, Modern Industry concludes that "the urge for security and the 
desire for company pensions is deep-seated and here to stay. Individual 
companies, whether their workers are organized or not, had better figure 
on facing up to this demand soon and either beat the gun or be ready for 
it when it comes." However, the suggestion is also made that "American 
workers have a wide streak of essential fairness. Most of them can be sold 
on a pension plan tied to company profits. And it's only a short second 
step to tie these profits upon which their pensions depend to heightened 
production" (59, p. 38). 

Progressive management has long recognized the importance of satisfy- 
ing workers' needs for security during old age, as is evident from the num- 
ber of retirement plans voluntarily undertaken and supported by manage- 


TABLE 24. Attitudes Toward Pensions in Relation to Other Benefits (after 

Modern Industry, 1950) 

Question: "If you had to choose between a pension at 65 or any one of the 
following benefits which would you rather have?" 


Prefer other Prefer both equally 

benefit, pension at 65, important, 

per cent per cent per cent 

Vacations with pay 25.0 42.2 32.8 

Sick leave with pay 26.5 38.8 34.7 

Life insurance 19.0 43.8 37.2 

Accident or health insurance 24.2 37.5 38.3 

Higher pay right now 31.8 37.0 31.2 

ment long before such plans became a major issue in union negotiations 
(11, 27, 38, 45). The interest of union leaders is attested to by the suc- 
cessful drives for pension plans in collective-bargaining negotiations or 
through the medium of strikes during recent years. Many issues remain 
to be resolved in the establishment and administration of pension plans, 
including that of the extent of employee contributions and the question 
of joint administration of pension plans (10, 17). Nevertheless, the in- 
creasing prevalence of pension plans shows an increasing recognition of 
the worker's need for long-range security. Simultaneously, such plans 
provide the medium for the satisfaction of this need. It may well be, as 
suggested by Becker, that "we are ... on the threshold of a new era in 
the development of a full measure of security for the workers of America 
and for their families" (17, p. 126) and that, in the years to come, we will 
have happier and more productive people because of steps taken to solve 
the problems of immediate and long-range security. 


1. American Legion, Employment Stabilization Service, To Make Jobs More Steady 
and To Make More Steady Jobs. St. Paul: Webb Publishing Co., 1942. 

2. Annual Wage and Guaranteed Employment Plans (Studies in Personnel Policy 
No. 76) New York: National Industrial Conference Board, 1946. 

3. E. W. Bakke, The Unemployed Man. Nisbet, Ltd., 1933. 

4. E. W. Bakke, Citizens Without Work. Yale University Press, 1941. 

5. S. Barkin, "Handling work assignment changes," Harvard Bus. Rev., 1947, 25, 

6. S. Barkin, "Human and social impact of technical changes" in Proceedings of 
Third Annual Meeting, Industrial Relations Research Assn., Champaign, 111.: In- 
dustrial Relations Research Assn., 1950, 112-27. 


7. S. Barkin, "The technical engineering services of an American trade union/' 
Int. Labor Rev., 1950, 61, 609-36. 

8. S. Barkin, "Trade-union attitudes and their effect upon productivity" in L. R. 
Tripp (Ed.), Industrial Productivity. Madison, Wis.: Industrial Relations Re- 
search Assn., 1951, 110-29. 

9. J. Barbash, Labor Unions in Action. Harper and Brothers, 1948. 

10. L. Boffo, Pension Plans in Collective Bargaining. University of Illinois, 1950. 

11. E. B. Brower, "Significant developments in pension plans," Conference Bd. Mgmt. 
Rec., 1948, 10, 277-79. 

12. R. Centers, "Motivational aspects of occupational stratification," /. Soc. Psychol., 
1948, 28, 187-217. 

13. S. N. F. Chant, "Measuring factors that make a job interesting," Pers. /., 1932, 
11, 1 ^. 

14. J. Chernick, Economic 'Effects of Steady Employment and Earnings. University 
of Minnesota Press, 1942. 

15. J. Chernick and G. C. Hellickson, Guaranteed Annual Wages. University of Min- 
nesota Press, 1945. 

16. Collectivist Ideology The Role of Job Dissatisfaction and Economic Ignorance 
(Executive Summary). Princeton: Opinion Research Corporation, 1947. 

17. "Crucial Issues in the Pension Problem" in Proceedings of the Second Annual 
Meeting (Part IV), Industrial Relations Research Assn. Champaign, 111., 1950. 
(See H. Becker, "Labor's approach to the retirement problem," pp. 116-26). 

18. N. M. Davis, "Attitudes to work among building operators," Occup. Psychol., 

19. P. F. Drucker, The New Society. Harper and Brothers, 1949, 82-84. 

20. S. S. Dunn, "A study of work attitudes," Bull. Industr. Psychol., Pers. Practices, 
1947, 3, 21-29. 

21. Economic Analyses of Guaranteed Wages (Bureau of Labor Statistics Bulletin 
No. 907), U.S. Govt. Printing Office, 1947. 

21a. Employment Stabilization. Industrial Relations Division, National Association 
of Manufacturers, 1952. 

22. C. E. Evans and La V. N. Laseau, My Job Contest (Personnel Psychology Mono- 
graph No. 1 ) . Washington: Personnel Psychology, Inc., 1950. 

23. Factors Affecting Employee Morale (Studies in Personnel Policy No. 85). New 
York: National Industrial Conference Board, 1947. 

24. H. Feldman, Stabilizing Jobs and Wages. Harper and Brothers, 1940, 

25. E. A. Grove and W. A. Kerr, "Specific evidence on origin of halo effect in measure- 
ment of employee morale," /. Soc. Psycho/., 1951, 34, 165-70. 

26. O. M. Hall, "Attitudes and Unemployment," Archiv. Psychol., No. 165, 1934. 

27. Handbook on Pensions (Studies in Personnel Policy No. 103). New York: Na- 
tional Industrial Conference Board, 1950. 

28. E. Hardin, H. G. Reif, and H. G. Heneman, Jr., "Stability of job preferences of 
department store employees," J. Appl. Psychol., 1951, 35, 256-59. 

29. A. Hoffman, "Worker and trade union attitudes toward technological change" in 
Proceedings 1950 Annual Fall Conference. New York: Society for Advancement 
of Management, 1951, 164-70. 

30. "Industry advised to prepare to grant guaranteed wages," Daily Labor Report, 
No. 188, (Sept. 26) 1951, A-2. 

31. C. E. Jurgensen, "Selected factors which influence job preferences," /. Appl. 
Psychol., 1947, 31, 553-63. 


32. C. E, Jurgensen, "What job applicants look for in a company/' Pers. PsychoL, 
1948, 1, 433-45. 

33. A. D. H. Kaplan, The Guarantee of Annual Wages. Washington: The Brookings 
Institution, 1947. 

34. D. Katz, "Good and bad practices in attitude surveys in industrial relations," 
Proceedings of Second Annual Meeting, Industrial Relations Research Association, 
Champaign, 111.: Industrial Relations Research Assn., 1949, 212-21. 

35. W. A. Kerr, "On the validity and reliability of the Job Satisfaction Tear Ballot/' 
/. Appl. Psychol, 1948, 32, 275-81. 

36. A. Kornhauser, "Attitudes of economic groups/' Pub. Opin. Quart., 1938, 2 

37. A. Kornhauser, "Psychological studies of employee attitudes/' /. Consult. PsychoL, 
1944,8,127-43. ' ' 

38. R. E. Larson, Insured Pension and Welfare Plans. University of Wisconsin School 
of Commerce, 1950. 

39. J. F. Lincoln, Incentive Management. Cleveland: Lincoln Electric Co., 1951. 

40. O. Lipmann, "Arbeitswissenschaft," An. Bet. und Arb., 1929, 3, 76-82. 

41. O. Lipmann, "Ergebnisse der Arbeitsforschung," An. Bet. und Arb. f 1929, 3 

42. O. Lipmann, "Ergebnisse der Arbeitsforschung," An. Bet. und Arb., 1930, 3, 

43. D. McGregor and I. Knickerbocker, Industrial Relations and National Defense: 
A Challenge to Management. Personnel, 1941, 18, 49-63. (Reprinted as M.I.T. 
Dept. of Economics and Social Science, Series 2, No. 7). 

44. R. N. McMurry, "The problem of resistance to change in industry." /. Abbl. 
PsychoL, 1947, 31, 589-93. 

45. "More on pensions/' Mgmt. Rec., 1950, 12, 217-21. 

46. A. Pigou, Employment and Equilibrium. The Macmillan Co., 1949. 

47. L. G. Reynolds and J. Shister, A Study of Job Satisfaction and Labor Mobility. 
Harper and Brothers, 1949. 

48. W. Reuther, "Labor's potential for a democratic world," Gen. Mag. Histor. Chron. 
(University of Pennsylvania), 1952, (Winter) 54, 88-96. 

49. A. M. Rose, Union Solidarity. University of Minnesota Press, 1952. 

50. E. Q. Rundquist and R. F. Sletto, Personality and the Depression. University of 
Minnesota Press, 1936. 

51. J. L. Snider, The Guarantee of Work and Wages, Harvard University Graduate 
School of Business Administration, 1947. 

52. }. Tiffin, "The uses and potentialities of attitude surveys in industrial relations," 
Proceedings of Second Annual Meeting, Industrial Relations Research Association, 
Champaign, 111.: Industrial Relations Research Assn., 1949, 204-211. 

53. L. A. Thompson, The Guaranteed Annual Wage and Other Proposals for Steady- 
ing the Worker's Income: Selected References. U.S. Govt. Printing Office, 1946. 

54. G. Watson, "Work Satisfaction" in G. W. Hartmann and T. Newcomb (Eds ) 
Industrial Conflict. The Cordon Co., 1939, 114-24. 

55. G. Watson, "Morale during unemployment," in G. Watson (Ed.), Civilian 
Morale. Harper and Brothers, 1942, 273-348. 

56. D. Weintraub, "Unemployment and increasing productivity" in Technological 
Trades and National Policy (75th Congress, 1st Session. House Document No 
360). U.S. Govt. Printing Office, 1937, 67-87. 


57. L. T. Wflkins, "Incentives and the young worker," Occup. Psychol, 1949, 23, 

58. L. T. Wilkins, "Incentives and the young male worker," Intern. J. Opin. Aft. Res., 
1950-51, 4, 540-61. 

59. "Workers sound off on pensions," Modern Industry, February 1 5, 1950, 19, 38-41 . 

60. S. Wyatt, and J. N. Langdon, Fatigue and Boredom in Repetitive Work (Industr. 
Health Res. Bd. Report No. 77). H.M. Stationery Office, 1938. 

61. Youth and the World of Work. Michigan State College, 1949. 

62. B. Zawadski and P. F. Lazarsfeld, "The psychological consequences of unemploy- 
ment," /. Soc. Psychol, 1936, 6, 224-51 (See also P. Eisenberg and P. F. Lazars- 
feld, "The psychological effects of unemployment," Psychol. Bull, 1938, 35, 

Status/ Recognition/ 

and the Role of Supervision 

"WHAT EVERY worker knows is this: that sooner or later 
the final joy of his work is settled, not by him or by his employer, but by 
the social standing awarded him by his fellow citizens" (51, p. 62). In 
this sentence Whiting Williams summarizes a highly significant con- 
clusion growing out of his experience as a worker on the railroads, in the 
mines and in the factories of this country and Europe (49, 50, 52). 

Status and Recognition 

As indicated 'in earlier chapters, in considering the sources of motiva- 
tion at work, Williams clearly recognized that the pay envelope and 
security are fundamental factors in determining the worker's attitude 
toward his job. The worker who walks to and from work for a week be- 
cause his child has unexpectedly worn out an extra pair of shoes, cannot 
help but be governed by the immediate compelling importance of regular 
and adequate pay. Living on the margin of bare subsistence, anything 
which will affect his earning capacity, either immediately or in the 
future, brings a rapid and certain response in the defense of the con- 
tinuity of his job. The narrowness of the "blanket" separating him from 
three meals a day and shelter and dependence goes far, writes Williams, 
in explaining the wary defense against anything which means curtailed 
rates or curtailed employment. It explains the lack of reality to the 
marginal worker of a bonus system which provides payment in the future 
instead of an immediate addition to salary; of any participation in 
management which does not favorably affect his pocketbook or help to 
regularize employment. 

However, as compelling as is this pressure for the Saturday night's 



earnings, this, according to Williams, constitutes only a prologue to the 
play of motives in industrial life. "It this mere first act were only the 
whole of the play, then the scientist could surely hope some day to dis- 
cover a satisfactory formula for wage revision" (51, p. 27). What makes 
this unlikely is that the pay check's ability to buy material things, par- 
ticularly when this is beyond the margin of the hunger minimum, is over- 
shadowed by the ability of the job and the check to buy an intangible 
something of equal importance and of vastly greater intricacy viz., his 
standing among his fellow workers. Chief among the sources of motives- 
in-work, writes Williams, is "the suprising vastness of the gap which every- 
where among the workers separates the holder of a 'swell 9 job from the 
holder of a 'bum 9 job, and most of all divides the possessor of ANY job at 
all from the luckless vagrant who possesses none and knows not where to 
find one" (51, p. 3). 

Throughout industry, according to Williams, there is found a system 
of caste based on work corresponding roughly to the more familiar hier- 
archy of caste based on property. The laborer transferred to the mill- 
wright gang, receiving only an additional two cents per hour, benefits 
from an immediate transformation in social status reflected not only in 
the congratulations of his fellow-workers but in the acceptance of his 
wife and family by those who had hitherto looked upon them as somewhat 
lower in the social level (51, p. 56). Everywhere among the workers, the 
nature of his job and not solely the earning power determines the social 
standing awarded the worker by his fellow-citizens. 


Attitude surveys have confirmed Williams' observations and have fur- 
ther shown that levels of job satisfaction and morale are related to the 
worker's perception of his status and role in the social microcosm of in- 
dustry and in society at large. Studies involving both young people and 
adults have shown that there is a wide range in the prestige values at- 
tached to various jobs. 1 

Illustrative findings from such surveys are presented in Table 25 which 
summarizes ratings of an identical list of 25 jobs obtained in studies con- 
ducted by Counts (7) in 1925, by Deeg and Paterson (9) in 1946 and 
by Welch (47) in 1948. The list used in the original study by Counts 
included 45 occupations. These were reduced to 25 in number in the 
follow-up study by Deeg and Paterson by deleting every other occupa- 

1 Representative investigations of occupational status, supplementary to those re- 
ferred to in this chapter, are described by Anderson (1), Baulder and Paterson (2), 
Byers (5), Duncan and Duncan (10), Form (13a), Hall (17), Hall and Jones (17a), 
Hyman (23), Lehman and Witty (30), Menger (32), Osgood and Stagner (34), 
Smith (39), Stevens (40), Stone and Paterson (41), Super (43), and Wilkinson (48) . 


tion as ranked by Counts, and then adding three of the original occupa- 
tions at widely separated points in the rank order of presentation. The 
follow-up study was undertaken because of the opinion, expressed by a 
school counselor, that the depression of the 1930's and World War II 
had "upset traditional irrational prestige values" as represented in the 
classification of jobs on the basis of social status. 

As is apparent from Table 25, rankings made by 475 subjects, includ- 
ing undergraduate and graduate students at the University of Minnesota 
and seniors attending high schools in Minneapolis and St. Paul, revealed 
extremely high consistency in the social ranking of occupations in spite 
of the lapse of 21 years and of the intervening economic and social changes 
in the United States. According to Deeg and Paterson, "the multitude 
of social, economic, and psychological factors determining the relative 
prestige of occupations has combined to operate in a consistent pattern" 
(9, p. 207). The uniformity of patterns in ranking occupations is fur- 
ther revealed in the study by Welch (47), involving 500 students of 
Indiana State College who ranked the same occupations in 1948 (Table 


The mechanisms through which such values are established are un- 
doubtedly complex. Among those involved are social perception, social 
norms, and variables contributing to the development of the ego, to the 
individual's perception of his status, and to role behavior, discussed in 
Chapter 5. Findings from a number of investigations, in particular by 
Counts (7), Nietz (33), Kay (27), Hartmann (18), and Super (42), 
support the view that the social prestige values attached to different oc- 
cupations reflect the predominant socio-economic conditions and stand- 
ards prevailing in a nation (9). Further evidence of this is found in a 
study by Davis (8) who, in 1927, investigated the attitudes towards oc- 
cupations of 112 public school children in the U.S.S.R. Occupations 
rated high in Counts' original list of 45 jobs included peasant, aviator 
and doctor, whereas banker, business man and minister were at the bot- 
tom of the list. Such findings, which place technical and manual jobs 
among the highly preferred occupations (also reported in studies by 
Russian investigators) are interpreted by Russian observers, following 
Stalin, as "a radical change in man's attitude toward labor, transforming 
labor from the shameful and heavy burden it was once considered to be 
into a thing of honor and glory/' Observations made in the U.S.S.R. in 
1935 by the author show that political indoctrination in the school sys- 
tem and in the vocational-guidance program was used extensively to 
increase the prestige value of occupations in which workers are needed, 



TABLE 25. Comparison of Social Status Ranks of Twenty-five Occupations 
Obtained in 1925, 1946, and 1949 (after Welch) 

Median rank orders 


1 2 3 

Counts Deeg and Paterson Welch 

1925 1946 1948 













Superintendent of schools 




Civil engineer 




Army captain , 




Foreign missionary 




Elementary-school teacher 












Traveling salesman 












Insurance agent 




Mail carrier 
























Truck driver 




Coal miner 








Hod carrier 




Ditch digger 




Note: Rho 1 and 2 = 0.97. Rho 2 and 3 = 0.98. 

and to turn young people from entering other occupations (28). Al- 
though similar forces may be operating in the United States, in the form 
of parental direction and pressure, it is nevertheless important to note 
that the absence of legal restrictions to job mobility such as exist in the 
U.S.S.R. and relative freedom in the operation of social forces in the 
United States add a quality of "honest opinion" to the social values at- 
tached to occupations by American children which does not appear to 
exist in parallel data applying to the population of the U.S.S.R. (28). 


Evidence that the social status accorded to the job helps determine 
the choice of occupations by young people is found in the study con* 


ducted in 1948 by the Social Research Service, Michigan State College 
(53). Students in the tenth and twelfth grades of Michigan high schools 
were asked to compare the occupations of teacher, farmer, telephone 
operator, foreman and supervisor, and assembly worker in a factory with 
others. In each case, the comparisons involved judgments as to the prestige 
value of the job in relation to ten others with which it is in competition 
in selecting careers or with which it is often associated. For example, the 
students were asked to select one alternative in the statement: "I think 
that the social standing or prestige of a telephone operator is higher than, 
equal to, or lower than that of a file clerk." 

For purposes of analysis, the proportion of students rating a job, e.g. 
telephone operator, "higher" than another in the list, was given an arbi- 
trary weight of 1; the proportion rating it "equal to" a weight of 2; and the 
proportion saying "lower" a weight of 3. Applying this method, as can be 
seen from Table 26, the prestige value of telephone operator among girls 
is found to be very slightly higher than that of file clerk. Such jobs as 
receptionist, stenographer, and bookkeeper are ranked above telephone 
operator, while lower prestige values are assigned to the jobs of sales 
clerk in a department store, elevator operator, etc. Other comparisons 
of the same sort lead to the conclusion that which of two or more jobs 
young people will accept depends a great deal on the relative prestige 
of these jobs. 


Employee-attitude surveys in industry also yield data on attitudes re- 
lated to job status from which inferences can be drawn concerning such 
motives as the needs for pride, prestige, social position, recognition, etc. 
Reference to Table 18 shows, for example, a relatively high per-cent men- 
tion for "work type" and "pride in important job" in letters submitted 
by General Motors Corporation employees in My Job Contest (12). 
Similarly, 18.5 per cent of employees in the six plants surveyed by the 
National Industrial Conference Board (13) in 1947 place "type of work" 
(Factor 71) among the five most important factors, and "type of work" 
is third in the list of factors when these are rated as first in importance 
(Table 18). If, as seems reasonable, opportunity for advancement 
(Factor 41) is viewed as bearing upon the satisfaction of prestige needs 
(remote as well as immediate), the data in Tables 17 and 18 can be in- 
terpreted as supporting the viewpoint that job characteristics or those 
company policies which enhance the standing of the job contribute to 
the satisfaction of basic needs. Among policies which are particularly 
pertinent in this connection are those applying to transfer and promo- 
tion, since although many employees may not be either capable or desir- 


TABLE 26. Prestige Level of Telephone Operator (from Youth and the 
World of Work, Michigan State College, 1949) 

Rank score 

Receptionist for a doctor 2.4 

Stenographer 2.3 

Bookkeeper 2.2 

Typist in an office 2 . 1 

File clerk 1.9 



Sales girl in department store 1 .6 
Clerk in grocery store 1 .6 

House-to-house salesman 1 .6 

Waitress in restaurant 1.5 

Elevator operator 1.5 

ous of moving on to another job, every employee likes to feel that he has 
the chance to "get ahead" ( 12) . 

Further evidence concerning the role of job status is found in the sur- 
veys conducted by the Survey Research Center, University of Michigan, 
described in Chapter 8. Among the dimensions of morale examined in 
the case of office employees of the Prudential Life Insurance Company 
(26) was intrinsic job satisfaction, characterized as the degree of satis- 
faction obtained by the individual employee from performing those tasks 
which constitute the content of the job. Questions used in obtaining an 
over-all index of intrinsic job satisfaction included 

1) "How well do you like the sort of work you are doing?" 

2 ) "Does your job give you a chance to do the things you feel that you 
do best?" 

(3) "Do you get any feeling of accomplishment from the work you are 

(4) "How do you feel about your work, does it rate as an important 
job with you?" 

The findings of the study showed that employees who were doing work 
of a more varied sort, or work requiring greater skill, had higher intrinsic 
job satisfaction than those working on less skilled jobs (24). However, 
in contrast to the results of earlier investigations by Katz and Hyman 
(25), and in conformity with those of Kornhauser and Sharp (29), the 
workers with higher intrinsic job satisfaction were not found to be the 
more productive workers. Moreover, they were not necessarily more 
identified with the company or more satisfied with the general policies of 
the organization. 


These and other findings of the study under discussion show that there 
is an intricate enmeshing of specific attitudes and specific motives. They 
also show, however, as do other investigations referred to in this section, 
that the status of the job itself its potential in arousing feelings of 
pride, worth and recognition is an important element in moulding at- 
titudes and in gratifying psychological needs. 

The Role of Supervision 

Experimental studies described in earlier chapters clearly indicate that 
the quality of supervision exercises a significant influence upon employee 
production, satisfaction and morale. Employee attitude surveys provide 
additional evidence that job satisfaction and morale are dependent upon 
the extent to which supervisors take into consideration employees 9 needs 
for recognition and status. 

Supervision and Morale: Data supporting this view are found in a study 
conducted during 1940 by the Florida Power and Light Company (38). 
Wide variations in morale from department to department in the 
company were revealed by the survey. Figure 29 presents the analysis of 
the morale situation in one of the departments of this company. The re- 
sults show clearly that the morale problems of this department were not 
centered around wages although, as the investigator points out, many 
people in the company assumed that "money tells the whole story of 
employee morale." Questions 23, 24, and 25 refer to wages. The attitudes 
of employees in this department towards wages are all "in the black/' 
that is above the corresponding company averages by the amounts 10.6, 
12.6, and 6.5, respectively. The largest deviation, in terms of unfavorable 
attitude, is with respect to Question 18, "Criticism in Public"; the value 
in this case is 18.2 points below the company average. Evidently the 
well-known principle of refraining from criticising employees in the 
presence of others had been violated flagrantly in this department 

Question 13, "Consideration and Courtesy Shown to Subordinates," 
reveals another source of unfavorable attitudes among employees in this 
department. In general, the survey revealed that in this department and, 
to some extent, in the company as a whole, the workers wanted more 
consideration, better treatment by the supervisory force. Such dissatis- 
faction as existed was not with the wage plan, but with the failure of the 
department head and his subordinates to recognize the workers' worth 
as human beings with the failure to satisfy the workers' needs for recog- 
nition, self-esteem and similar motives. 

Similar effects, bearing on absenteeism, have been noted in a study 
conducted in a large public utility by the Survey Research Center, Uni- 
versity of Michigan (25a). Of male office workers who had an average 


H 8 



pf lf:li?|iW|M^ 

.l3pldSS#.i ^SiBfrlfiJ.* 


G ^ 

D* SI * 


^r |ss r " 

C-S& CJ """KC53JW 

^ S3 " guilder, 

^! CU bfi to 



o K -aa"l.ra ?5*^- 



JS % 













absence rate of 1 day in 6 months, 69 per cent stated "I feel free to discuss 
important things about my job with my supervisor" The proportion 
giving the same reply among those with an average absence of 4 or more 
days in 6 months was 29 per cent. Although, as stressed by the investi- 
gators, many aspects of the social organization of the plant must be 
considered in evaluating such findings, the facts cited above serve to 
illustrate again the important role of the supervisor in promoting atti- 
tudes associated with desired behavior on the part of employees. 

Illustrations from Other Surveys: The findings of many employee- 
attitude surveys agree in showing that the quality of supervision con- 
stitutes a major factor in determining the level of employee morale. 
According to Houser (20), the differences in departmental average 
morale scores shown in Figure 21 reflect primarily differences among sub- 
executives in "winning the favor of employees/' 

A study by Kornhauser and Sharp (29), conducted in the Badger- 
Globe Mill of the Kimberly-Clark Corporation, showed "favorable" 
attitudes in the case of 29 per cent of employees in Department A. In 
Department B, engaged in the same kind of work and characterized by 
similar working conditions, 71 per cent of the employees had "favorable" 
attitudes. An intimate study of the attitudes and conditions of the two 
groups left no doubt that the one outstanding causal influence was the 
unfortunate nature of supervision in Department A. 

Further evidence of the influence of supervision in determining morale 
and in tapping motives is found in the work of Bergen (3) who used a 
questionnaire involving a combination of an attitude scale with questions 
of the multiple-choice type in measuring both over-all morale and reac- 
tions to particular policies of 1,000 employees from selected offices and 
factory departments of a manufacturing company. Results showed a 
range in average departmental morale scores from 45.9 to 69.4 with an 
average of 57.1 for all employees. A further analysis revealed that these 
variations in morale, although also influenced by other factors, were due 
largely to differences in supervision and leadership. 

Quality of Supervision and Job Satisfaction: Supervisory attitudes and 
conduct are determinants of job satisfaction as well as of morale. Evidence 
of this is found in investigations (35, 36) by the Survey Research Center, 
involving 8,000 nonsupervisory employees and 750 of their immediate 
supervisors in a large public utility. Relationships were found between 
employees' satisfactions and such supervisory variables as degree to which 
supervisor is concerned with employees as individuals; type of recognition 
given by supervisor for good work; extent of supervisory encouragement of 
employee participation in decision-making; degree to which supervisor 
takes sides with employee or with management in conflicts between 


them, etc. According to Elliott (11), when an employee describes the 
"kind of boss" he wants, he emphasizes the need for frequent praise and 
constructive criticism, personal interest, listening to ideas about the job, 
telling about changes before making them, and personal faith and con- 
fidence in the worker. Such findings confirm the conclusion, reached 
earlier by Hoppock (19), from an investigation of job satisfaction, that 
the employee wants a fair-minded boss who makes an effort to understand 
his problems, to answer his questions, to give consideration to his needs. 
Additional support for this view is found in a study of work adjustment 
in relation to family background by Friend and Haggard. One conclusion 
from this investigation is that job satisfaction, "derived from certain 
subtle rewards which dovetailed with the individual's personal needs, 
can be traced back to attitudes of supervisors . . . which fostered the 
individual's need of worth whileness" (14, p. 12). This and similar effects 
occur, in part, because of the dependence of the employee upon his 
supervisor. Throughout industry, as Gardner points out, "... everyone 
seems to be looking upward with his attention focused upon the people 
above him and especially upon his boss . . . Each subordinate is con- 
cerned over just how his boss feels about him. 2 He wonders if his work is 
satisfactory, if he makes a good appearance, if his boss thinks he talks too 
much or not enough, or if he knows just what his boss does expect'' (15, 
p. 9). 


Studies described above show clearly that the worker looks to his super- 
visor for the satisfaction of his desires for approval and recognition; of 
the need for self-respect, and of other ego-involved needs. His status in 
the plant, as he sees it, and as others see it, appears to be in considerable 
part a function of the quality of supervision. The importance of work 
status achieved through recognition of human needs by supervision and 
in other ways is enhanced by the universal connection between a man's 
work status and his community standing. According to Williams 

". . . this it is that constitutes by far the most important as well as the 
most elusive of all the Conditions' of work. This is what enlarges enor- 
mously the boundaries of a man's job. This is what provides those 'social 
handles' of the pay-cup that 'social fringe' which makes a man's work 
infinitely more than that combination of mud, shovel, and arms which 
so fills the eye of the observer. Away out beyond these extends the 
thought of the doer of the task. Constantly he pictures the reward which 
is to follow. Of that reward only a part is put into the pay envelope. The 
rest of it spills over into satisfactions astonishingly intricate and im- 

* Not italicized in the original. 


ponderable and indispensable. Among these, however, the assignment 
of a gratifying social status is only the final capstone set upon a whole 
series of lesser but ascending rewards" (51, p. 58). 

A practical fact that cannot be overlooked is the difficulty of enlarging 
the status of workers employed in the many relatively unskilled or repeti- 
tive jobs which characterize work in the "big factory/' One source of diffi- 
culty here arises from social norms 8 existing in the community outside of 
the plant. Thus, as is pointed out by the president of the Lincoln Manu- 
facturing Company, certain kinds of work, e.g. housework, are held to 
be "degrading" if done for pay; highly respected, if done by ourselves in 
our own homes. Driving one's own car is a source of prestige; acting as 
a chauffeur for pay may be looked upon as a menial task. A "white collar" 
job carries prestige in the community as well as in the plant. By contrast, 
manual work particularly unskilled labor is accorded little if any social 
recognition (31, p. 93). 


Such views create difficulties in satisfying the feeling among workers, 
noted in a British study, that "a job should have a significance beyond 
satisfying immediate needs;" the feeling that the job is both an essential 
part of the factory work and "important to the country and to the con- 
sumer" (46a, p. 191). Nevertheless, while it is true that low job status 
is, in part, a problem arising from the "views" of society at large, there is 
still much that can be done by industry to enhance the status and enrich 
the role of workers on even the lower rated jobs. Thus, job enlargement 
and job rotation on respective tasks may not only serve to reduce the 
feeling of boredom (37, 44, 45, 46) but simultaneously raise the status of 
the worker "in his own eyes/' in those of his fellow workers, and in the 
opinion of his friends and associates, his home community. A prime effect 
of the opportunity for participation in decision-making may well be to 
raise the status of the worker as he sees it and as it is viewed by other peo- 
ple in the plant. 4 The worker, as do people in other situations, wants to be 
in the role of an "important person." He seeks to enjoy this feeling of 
worth the recognition and respect of others. As has been suggested by 
Burns (4), the worker's point of view implies that a railwayman, an 
engineer, a machinist, an oiler, is bound, just as "intellectual" workers, 
by the honor of a calling. There is pride in it pride in the work one does 
and in its necessity. To "count," to be considered worthwhile, because of 
his job, represents a basic need of the industrial worker. To understand 

See pages 93-98. 
4 See pages 9&-103. 


the strength of this motive, the force of its interplay with encouraging 
approvals and opposing disapprovals, is one road toward a fuller ap- 
preciation of the manifold sources of motivation and morale in industry. 


1. W. A. Anderson, "Occupational attitudes of college men," J. Soc. Psycho/., 1934, 
5, 435-466. 

2. L. Baulder, and D. G. Paterson, "Social status of women's occupations," Occupa- 
tions, 1948, 26, 421-424. 

3. H. B. Bergen, "Finding out what employees arc thinking," Conference Board 
Mgmt. Rec., (April) 1939, 53-58. 

4. C. D. Burns, The Philosophy of Labor. Allen and Unwin, Ltd., 1925. 

5. B. H. Byers, "How the G.I. rates the job," The Nation's Schools, 1946, 37, 51. 

6. R. Centers, "Motivational aspects of occupational stratification," J. Soc. Psychol., 
1948, 28, 187-217. 

7. C. S. Counts, "Social status of occupations," Sch. Rev., 1925, 33, 16-27. 

8. J. Davis, "Testing the social attitudes of children in the government schools in 
Russia," Amer. J. Soc., 1927, 32, 947-952. 

9. M. E. Deeg and D. C. Paterson, "Changes in social status of occupations," Oc- 
cupations, 1947, 25, 205-08. 

10. H. G. Duncan and W. L. Duncan, "Attitudes of college students toward profes- 
sions," /. Educ. SotioL, 1935, 35, 454-461. 

11. T. A. Elliott, "What bosses and workers want," Pers. /., 1950, 28, 372-73. 

12. C. E. Evans and La V. N. Laseau, My Job Contest (Personnel Psychology Mono- 
graph No. 1). Washington: Personnel Psychology Inc., 1950. 

13. Factors Affecting Employee Morale (Studies in Personnel Policy No. 85). New 
York: National Industrial Conference Board, 1947. 

13a. W. H. Form, "Toward an occupational social psychology," /. Soc. Psychol., 
1946, 24 (First Half), 85-99. 

14. J. G. Friend and E. A. Haggard, "Work adjustment in relation to family back- 
ground," Appl. Psychol. Monog., No. 16, 1948. 

15. B. B. Gardner, Human Relations in Industry. R. D. Irwin, 1946. 

16. V. Gsovski, Ekments of Soviet Labor Law (Bureau of Labor Statistics Bulletin 
No. 1026). U.S. Govt. Printing Office, 1951. 

17. G. W. Hall, "Social prestige values of a selected group of occupations," Psychol. 
Bull, 1938, 35, 696. 

17a. J. Hall and D. C. Jones, "Social grading of occupations," Brit. J. Sociol, 1950, 
1, 31-55. 

18. G. W. Hartmann, "Prestige of occupations," Pers. /., 1934, 13, 144-152. 

19. R. Hoppock, Job Satisfaction. Harper and Brothers, 1935. 

20. J. D. Houser, What the Employer Thinks. Harvard University Press, 1927. 

21. "How to find out what your workers think about you," Factory Mgmt. Mainten., 
(August) 1948, 106, 81-91. 

22. R. L. Hull, "Measuring employee attitudes a proving ground for personnel 
policy and practices," Conference Board Mgmt. Rec., (November) 1939, 165-72. 

23. H. H. Hyman, "The psychology of status," Archives of Psychology, No. 269, 1942. 


24. D. Katz, ''Morale and motivation in industry" in Current Trends in Industrial 
Psychology. University of Pittsburgh Press, 1949, 145-71. 

25. D. Katz and H. Hyman, "Morale in war industry" in T. Newcomb and E. Hartley 
(Eds. ) , Readings in Social Psychology. Henry Holt & Co., 1947, 437-47. 

25a. D. Katz and R. L. Kahn, Some Recent Findings in Human Relations Research 
(Mimeographed Report). Survey Research Center, University of Michigan, 1952. 

26. D. Katz, N. Maccoby and N. C. Morse, Productivity, Supervision and Morale in 
an Office Situation (Human Relations Series 1, Report 1 ) . University of Michigan, 

27. L. W. Kay, "The relation of personal frames of reference to social judgments " 
Arch. PsychoL, No. 183, 1943. 

28. F. J. Keller and M. S. Viteles, Vocational Guidance Throughout the World. New 
York: W. W. Norton & Company, Inc., 1937, Chapter 6. 

29. A. W. Kornhauser and A. A. Sharp, "Employee attitudes/' Pers. /., 1932, 10 

30. H. C. Lehman and P. A. Witty, "Further study of the social status of occupa- 
tions/' /. Educ. Sociol, 1931, 5, 101-112. 

31. J. F. Lincoln, Incentive Management. Cleveland: Lincoln Electric Co., 1951. 

32. C. Menger, "The social status of occupations for women/' Teachers College 
Record, 1932, 33, 696-704. 

33. J. A. Nietz, "The depression and the social status of occupations," Elem. Sch. /., 
1934, 35, 454-461. 

34. C. E. Osgood and R. Stagner, "Analysis of a prestige frame of reference by a 
gradient technique," J. Appl. PsychoL, 1941, 25, 275-290. 

35. D. C. Peltz, Relation of Six Supervisory Traits to Nine Employee Attitudes. Un- 
published study, 1950, referred to in E. A. Fleishman, Leadership Climate and 
Supervisory Behavior. The Ohio State University, 1951. 

36. D. C. Pelz, "The effect of supervisory attitudes and practices on employee satis- 
factions," Amer. PsychoL, 1949, 4, 283-84. 

37. S. A. Raube, "The problem of boredom," Conference Ed. Mgmt. Rec. t (De- 
cember) 1948, 10, 565-75. 

38. McG. Smith, "Mending our weakest links," Adv. Mgmt., 1942, 7, 77-83. 

39. M. Smith, "An empirical scale of prestige status of occupations." Amer. Sociol. 
Rev., 1943, 8, 185-192. 

40. R. B. Stevens, "The attitudes of college women toward women's occupations/' 
/. Appl. PsychoL, 1940, 24, 615-627. 

41. C. H. Stone and D. G. Patcrson, "Dissatisfaction with life work among adult 
workers," Occupations, 1942, 21, 219-221. 

42. D. E. Super, "Occupational level and job satisfaction," J. Appl. PsychoL. 1939, 

43. D. E. Super, The Dynamics of Vocational Adjustment. Harper and Brothers, 1942. 

44. M. S. Viteles, "Man and machine relationship the problems of boredom" in 
Proceeding 1950 Annual Fall Conference. New York: Society for Advancement 
of Management, 1951, 129-38. 

45. C. R. Walker, "The problem of the repetitive job," Harvard Bus. Rev. t 1950, 28, 
3, 54-58. 

46. C. R. Walker and H. R. Guest, 'The man on the assembly line/' Harvard Bus. 
Rev., 1952, 30, 3, 71-83. 

46a. J. Walker and R. Marriott, "A study of some attitudes to factory work," Occup. 
PsychoL, 1951, 25, 181-91. 


47. M. K. Welch, "The ranking of occupations on the basis of social status," Occu- 
pations, 1949, 27, 237-41. 

48. F. Wilkinson, "Social distance between occupations," Social, and Soc. Res., 1929, 

49. W. Williams, Full Up and Fed-UpThe Worker's Mind in Crowded Britain. 
Charles Scribner's Sons, 1921. 

50. W. Williams, Horny Hands and Hampered Elbows. The Worker's Mind in 
Western Europe. Charles Scribner's Sons, 1922. 

51. W. Williams, Mainsprings of Men. Charles Scribner's Sons. 1925. 

52. W. Williams, What's on the Worker's Mind by One Who Put on Overalls to 
Find Out. Charles Scribner's Sons, 1920. 

53. Youth and the World of Work. Michigan State College, Social Research Service, 


Attitudes toward the Union 

ACCORDING to estimates based on union membership 
reports and claims, union membership in the United States has increased, 
as shown in Figure 30, from approximately 3 million in 1930 to between 
14 and 16 million in 1950. Union membership in the latter year included 
approximately one- fourth of the total labor force (32, p. 113). 

The rise of unionism in the United States has been ascribed to the 
success of labor organizations in satisfying workers' needs. According to 
Golden and Ruttenberg, spokesmen for organized labor, unions have con- 
tributed to the satisfaction of three major categories of human needs, viz. 


10- - 

Trend in Union Membership 

1900 1910 1920 1930 1940 1950 


Botod on Union fttportt 

Figure 30.* 

* From 50 Years' Progress of American Labor, Monthly Labor Review The 
35th Anniversary Issue, (July) 1950, 71, No. 1. 


"1. Economic an adequate plane of living and necessary amount of 
'wage and job protection; 

"2. Psychological the personality needs of freedom of action, self- 
expression and creative outlets; 

"3. Social the ties and bonds of group relations and community life" 
(13, p. 7). 

In a discussion of UAW and Walter Reuther, such claims have been 
enlarged into the assertion that "It [the Union] alone has brought a sense 
of human worth into an area dominated by robots, pistons, and dollars 
and that, more than anything else is the measure of its triumph" (16, 
p. 44). 

Consideration of union functions have led a number of social scientists 
to the even broader conclusion that the union represents the prime me- 
dium for furnishing "substitute" satisfactions for employee needs and 
demands which are frustrated by the job conditions under which people 
work in modern industry (24) . Thus, according to Krech and Crutchfield 

"While it is true that the individual worker may find substitute satis- 
factions in many different kinds of social organizations in fraternal, 
religious, or other types it is highly probable that he will seek many of 
them in labor unions. There are .various reasons for this. Under present 
conditions almost every worker is familiar with labor unions, comes into 
contact with them, and sometimes is even forced to join them. On the 
other hand, the opportunity and compulsion to become acquainted with 
or join other types of social organizations is not so great. Second, the labor 
union, by and large, can better meet most of the worker's needs and de- 
mands than can other organizations. 1 As we have seen . . . most social 
organizations will generally reflect the major needs of its members, and 
labor unions will therefore be more 'tailored' to the needs of workers 
than will religious organizations or other less homogeneously composed 
social organizations" (19, p. 547). 

Reasons cited by Krech and Crutchfield for such efforts (representing 
also reasons other than the "economic" drive for joining unions) are as 

"1. Labor unions frequently provide the individuals with a sense of 
participation in deciding vital issues . . . 

"2. Labor unions frequently provide the individual with an opportunity 
to express himself in speech and action with much more freedom from 
fear of reprisal than is possible on a job. . . . 

"3. Labor organizations provide the worker with opportunities to 
achieve positions of leadership and authority opportunities that are 
denied him on the job" (19, p. 548). 

Unions, it is added, counteract tensions and frustrations produced by 
job conditions and help satisfy needs and wants by sponsoring many other 

1 Not italicized in the original. 


activities. A sense of economic security is provided through the establish- 
ment of credit unions and banks. Other needs are satisfied through educa- 
tional, recreational, and cultural programs which "can enrich a life made 
monotonous by the assembly line." The desire for self-expression is pre- 
sumably better achieved because, "through the political programs of 
many unions, he [the worker] can participate in local, state, and national 
politics in a more personal and genuine sense than would otherwise be 
possible. The modern labor union," write Krech and Crutchfield, "like 
the modern drugstore, caters to almost every need of the public" ( 19, p. 

The viewpoint that the worker looks to the union for the satisfaction 
of such needs has been challenged by management representatives. Dur- 
ing the intensive drive toward unionism in the late 1930's and early 1940's 
many executives contended that "workers don't want to join unions, but 
they are being forced into them by racketeering labor agitators supported 
by self-seeking* politicians" (5, p. 118). More recently, many have ac- 
cepted the view that management's abuse of its power, particularly during 
a period of lowered job mobility in the depression of the 1930's, con- 
tributed to the growth of unionism in the United States (14) . However, 
there are still many sincere and thoughtful observers who question the 
position that the increases in union membership represent a spontaneous 
expression of solidarity on the part of a universally exploited and deeply 
dissatisfied class of the population (31), and also the further view that 
the union is required to overcome the frustrations induced by the "big 
factory" and the Machine Age. There is also a disagreement with the view 
that management, because it has the "attitudes, habits, and values of 
middle-class groups" cannot understand what the worker wants (6) and 
must turn to the union and its representatives for enlightenment. 

The existence of such diverse views, reflecting perhaps the varying so- 
cial norms of the interested groups (22), points to the need for a con- 
sideration of such objective data on workers' attitudes toward unions as 
are available from studies in this field. An examination of attitude surveys 
from this point of view can also provide further clues as to the basic 
motives which require particular consideration in dealing with problems 
of motivation and morale in industry. 


Data bearing on the question as to whether youth wants to belong to 
labor unions is found in the study conducted by the Social Research 
Service, Michigan State College, in 1948. As shown in Table 27, the great 
majority of Michigan High School students feel that they should have 


TABLE 27. The Attitudes of Young People Toward Union Membership (from 
Youth and the World of Work, Michigan State College, 1949) 

1 would prefer a job with a concern where the workers: 

Per cent 

Must be members of a labor union 9 

Can join a labor union if they want to 59 
Have no labor union 17 

Don't know which I would prefer 12 

No response 3 

Total 100 

an opportunity to join a union if they so desire. However, less than 10 
per cent want compulsory membership, and 17 per cent want nothing to 
do with unions. On the other hand, as shown in Figure 31, "children of 
manual workers, who are more likely to have had actual experience with 
unions, more often favor compulsory union membership or the union 
shop/' Conversely, "children of proprietors, managers, officials, profes- 
sional and clerical workers more often want jobs where there is no union" 
(33, p. 13). 

The Influence of Socio-economic Status: Further data showing the 
effect of socio-economic status and union affiliations of parents upon atti- 
tudes of younger people are found in studies by Weschler (27) who 
used an "error-choice" method (15) in evaluating college students' atti- 
tudes toward labor and management. Results showed that, in a college 
population, the subject's income provided a significant clue to his atti- 
tudes toward labor or management, since those in economically poorer 
brackets tended to score significantly higher on a "pro-labor" continuum 
(Average Score, 16.00) than did individuals in higher income brackets 
(Average Score, 13.44). 2 Furthermore, those individuals whose parents 
belonged to labor unions scored higher (Average Score, 16.36) on the 
same continuum than did those whose parents belonged to management 
groups (Average Score, 13.00). 

Striking evidence of the influences of socio-economic status and paren- 
tal affiliation with a union on the attitudes of young people toward "labor" 
is found in a study by Centers (3). Subjects were approximately 1,000 
students, representing the entire high-school population of a town of 
30,000 people situated approximately 35 miles from the Manhattan 
section of New York City. During a single day near the end of the 
semester these students at the school were asked by their teachers to 

2 Difference between average scores significant at or below the 5 per cent level. 



Occupation of 



Skilled Workers 
and Foremen 

Sales and 
Kindred Workers 


Managers, and 












Prefer a Job Where Workers: 

Have no 

Must join 
a union 


Can join a 

union if 
they want to 

Don't know 
or no opinion 

Figure 31. Boys' Opinions about Union Membership, by Occupation of 
Father * 

answer anonymously a brief questionnaire labeled A Survey of Student 
Opinions on Some Public Issues and Policies, which included eight 
questions relating to "labor." 

Test-retest reliability of the labor scale proved to be .98. The scale 
proved to have a substantial degree of validity in the sense that responses 
clearly differentiated members of a Socialist Club, on the one hand, and 
members of the Society for the Advancement of Management, on the 
other. In fact, all members of the Socialist Club (N = 25) were rated 
as Very Pro-labor on the scale; none of the members of the Society for the 
Advancement of Management (N = 20) fell into this category, and all 
but six were rated as Very Anti-labor. 

* By permission, from Youth and the World of Work, Michigan State College, 1949. 



As can be seen from Table 28, which represents findings from the ad- 
ministration of the scale to the high-school population, the frequency 
of "pro-labor" attitudes is high among the children of semiskilled and 
skilled manual workers and low among the children of executives and 
professional people. Conversely, "anti-labor" attitudes are infrequent 
among the children of the "working" population. Systematic breakdowns 
of the data showed that at the intermediate occupational strata, i.e. white 
collar and skilled manual workers, differences in attitudes were associated 
with labor-union membership on the part of the parents. However, accord- 
ing to Centers, "these responses, although influenced by certain . . . 
other variables to some degree, were more strongly related to parental 
occupation than to any of these others" ( 3, p. 322) . 

TABLE 28, Labor Attitudes of Adolescents (after Centers) 



Per cent 

For "very 
are signifi- 
cant between 







1. Business executive 
2. Professional 
3. Small business 
4. White collar 
5. Skilled manual 
6. Semiskilled manual 
7. Unskilled manual 











2 & 3, 4, 5, 6 


* Differences not tested for significance because of the very small number of cases 
in the samples. 


N.I.C.B. 1947 Survey: Facts concerning the measured attitudes of 
employed workers toward unions are available from a number of studies. 
The large-scale surveys described in Chapter 13 provide data from which 
indirect inferences might be drawn concerning the importance attached 
to union membership by workers. Thus, reference to Table 17 shows that 
only .9 per cent of employees in the six plants surveyed by the National 
Industrial Conference Board in 1947 referred to "labor unions" (Factor 
34) as first in importance among 71 morale factors, and this factor was 
rated among the jive most important factors (Table 18) by only 5.8 per 
cent of the employees involved (11). The factor "labor union" occupies 
approximately the 25th position in a percentage ranking of factors selected 


as first in importance by employees. It is of interest to note, in passing, 
that ^abor unions" does not appear at all in the list of factors predicted 
by employers to be first in importance in the opinion of their employees. 
On the other hand, "labor unions" occupies the third position in a rank- 
ing of the 71 factors by 42 labor-union leaders as representing what they 
believed to be first in importance in the opinion of union members. 

CMC My Job Contest: Of further interest in considering this possi- 
bility is the fact that the investigators responsible for coding letters written 
for My Job Contest, by employees in the highly unionized General Motors 
Corporation plants, found it unnecessary to use "labor unions" in analyz- 
ing and presenting the findings of this survey. The final list of 58 themes 
shown in Table 18 was derived from an original list of 77 themes, from 
which were eliminated those which proved to lack "sensitivity" in report- 
ing results. Originally, five items had been introduced experimentally 
into the coding structure to reflect the entrants' comments about the 
union. References to these items totaled to less than one-per-cent men- 
tion, and they were not treated as separate themes in the final tabulations. 

The General Motors Corporation contest also provided an opportunity 
for contestants to write in comments under a large "PS" on the reverse 
side of the M/C entry blank. Approximately 7.2 per cent of the entrants 
made "PS Comments." Of the total comments entered, only 1.4 per cent 
referred to unions in comparison, for example, with 22.8 per cent referring 
to "methods, tools, upkeep"; 13.3 per cent to "working conditions"; 6.8 
per cent to "training and education," etc. However, the paucity of refer- 
ences to unions by M/C contestants is interpreted by Evans and Laseau 
as "establishing the point clearly that a management sponsored survey 
should not be expected to elicit comments about the union, especially 
in the light of the free-response condition under which the entrant was 
operating" (9, p. 44). 

Job Satisfaction of Railroad Workers: The role of the union as a source 
of satisfaction on the job has been investigated by Stagner in a study of 
railroad workers (26a), conducted with the co-operation of both man- 
agement and union officials. This included a sample of 715 drawn from 
a total of about 22,000 employees, not including maintenance-of-way 
gangs. A job satisfaction score was computed for 100 respondents, chosen 
at random, by weighting 15 items of the questionnaire used in the survey. 
Of these 15 items, that pertaining to "union-management relations" was 
found to be the most discriminating in differentiating between the 50 
more satisfied and the 50 less satisfied members of this group. According 
to Stagner, this finding, and the further fact that "grievance handling" 
is the second most discriminating item, seems "plausible in a highly 
unionized industry" (26a, p. 298), 


Why Workers Join or Refrain from 
Joining Labor Unions 

The surveys reported so far deal only with over-all attitudes toward 
unions and labor. A number of surveys have concerned themselves with 
the question of 'why -workers join or jail to join labor unions. 


Illustrative of investigations bearing upon reasons for joining unions 
is a study by Chamberlin (5) who, in 1935, interviewed 200 men em- 
ployed in textile mills, in Massachusetts, including 100 union members 
and 100 nonunion workers. Their answer to his inquiries indicated that 
90 per cent of union members and only 38 per cent of nonunion em- 
ployees believed that the unions get results. To a request for reasons for 
which they would join the union, nonunion employees gave the following 
in the order noted: (1) because fellow workers had joined 9 , (2) a feeling 
of greater security; (3) because a union is the only way that the working 
man can get results; (4) a liking for such organizations. The principal 
objection of nonunion men to unions was the union's failure to get results 
(45 per cent), with the type of leader running a close second (41 per 


Data on the relationship between attitudes toward the union and 
job satisfaction are found in the study of the "man on the assembly 
line/' by Walker and Guest (40), described on page 294. In the course 
of interviews, workers were asked to tell what they thought about the 
union and why. An evaluation of responses by the investigators showed 
that 66 per cent were generally favorable to the union, which had been 
organized shortly after plant operations began, in 1949. From these find- 
ings, and a further consideration of the nature of the qualitative remarks, 
it is concluded that "a majority of the men favored the union, and that 
its role at Plant X was an important one'' (40, p. 132). However, in an 
analysis of the relative importance of the union in comparison with other 
job conditions rated by employees, it was found that only 2 in the sample 
of 180 employees gave the "union" as their first reason for liking their 
job at Plant X, in contrast with the large number who gave "good pay" 
or "steady work" as the "most important reason for liking the job." 
Nevertheless, the investigators' examination of the "usual reasons" why 
workers join or continue to support a union leads to the opinion that 
wages, job security, hours, etc. are not applicable explanations at Plant 
X. Instead, it is suggested that the union served to counterbalance a lack 
of personal satisfaction with immediate work experience- that "the 


union met in part the psychological and social needs which work in the 
plant had created/' The union, it is added, represents an "emotional 
as well as economic dimension in the workers' attitudes ... a kind of 
psychological bulwark against pace and boredom and against the bigness 
and impersonality of management" (40, p. 133). 


Questions pertaining to reasons for joining or not joining the union 
were included in a survey sponsored by the National Association of Manu- 
facturers and conducted by Cherrington and Roper. This involved inter- 
views with approximately 6,000 employees representing a cross-section 
of the working population of the United States. Replies to the question 
"Why do you think most union members join a union?" are summarized 
in Table 29. While showing that "more money/' "benefits" and "bargain- 
ing power" are impelling forces in attracting employees to unions, the 
findings are also interpreted as showing that a very small proportion of 
workers join unions primarily for job security and that "they evidently 
look to management for job security" (30, p. 8). Answers to the question 
"Why do you think most nonunion men do not join a union?" also sum- 
marized in Table 29, give prominence to "fear" of the employer as a reason 
for refraining from joining unions. However, the combination of a pro- 
portion giving the answers "satisfied with conditions," "don't believe in 
unions," and "mistrust unions and union leaders" shows, according to the 
National Association of Manufacturers, that "the reasons for not joining 
[unions] are considerably more on the side of what the unions don't offer 
than of the fear complex many have about the employer" (30, p. 9) . 

TABLE 29. Reasons for Joining or Not Joining Unions National Association 
of Manufacturers Survey (Undated) 

Question: Why Do You Think Most Union Members Join a Union? 

Reasons Per cent 

More money 27.5 

Benefits they get 24.9 

Bargaining power 18.9 

Job security 3.0 

Question: Why Do You Think Most Nonunion Men Do Not Join a Union? 

Reasons Per cent 
Fear employer 25.7 

Satisfied with conditions 19.7 

Don't believe in unions 16.7 

Distrust unions or union leaders 12.9 


In a survey conducted by the Opinion Research Corporation (20), in 
1946, a cross-section of the general public (N = 3,320) and of manual 
workers composed of workers from the general public plus a supplemen- 
tary sample of 882 cases (N = 1,801), was asked "When a man joins a 
union, is it more often because he wants to, or because the union compels 
him to?" To this question 63 per cent of the general public, 55 per cent 
of nonunion manual workers, and 41 per cent of union manual workers 
replied "union compelled/' Only 47 per cent of union manual workers 
replied "wants to," and the percentage of nonunion manual workers giv- 
ing this reply, like that of the general public, was found to be in the neigh- 
borhood of 25 per cent. 

More detailed data on reasons for joining or not joining unions are 
found in a study of, workers' reactions to industrial problems in a war 
economy conducted, in 1942, by Link (23) . This involved interviews with 
a nation-wide sampling of a thousand individuals, consisting mainly of 
industrial workers and their wives. Of the interviews conducted, 70 per 
cent were in homes of industrial workers and 30 per cent in "white-collar" 
homes of comparable economic status. The distribution of union mem- 
bership in these homes is shown in Table 30. Half of the interviews were 
with men and the other half with women, mostly the wives of workers. 

Among the questions asked were ( 1 ) "Do you think it would be a good 
thing if ALL WORKERS were required to join or belong to a union?" and (2) 
"Do you think that workers should be forced to STAY IN a union if they 
want to resign or get out?" Of individuals with union backgrounds 61 per 
cent believed that all workers should be required to join a union. Of 
individuals without union backgrounds in the home 61 per cent believed 
that all workers should not be required to join a union. At the same time, 
a large proportion of all industrial workers and their wives were opposed 
to compulsory maintenance of union membership, as evidenced in the 
fact that 84 per cent of those without union backgrounds and also 72 per 
cent with union backgrounds answered "No" to the question "Should 
workers be forced to stay in a union?" 

Table 31 shows the reasons given by 387 individuals in the nation-wide 
sampling who said that all workers should be required to join a union and 
of 495 who said they should not. The reason most frequently given for 

TABLE 30. Frequency of Union Membership in Homes Surveyed in Links 
Study (Per Cent) (after Link, 1942) 

Industrial Other 

homes homes Totals 

Union members, per cent 33.0 3.5 36.5 

Nonunion, per cent 36.5 27.0 63.5 

Total numbers 695 305 1000 



compulsory maintenance of membership by 26 individuals of 132 fa- 
voring union maintenance was "It would weaken the union if they 
got out." By contrast, 381 individuals in the sample of 1,000 objected to 
compulsory maintenance of membership on the grounds that the worker 

TABLE 31. Opinions as to Why Workers Should or Should Not Join a Union, 
Per cent (see note) (after Link, 1942) 

Part 1: Reasons Why All Workers Should Be Required to Join or Belong 
to a Union 

By membership 


Total homes Nonunion 
Secure better working conditions, get better 

wages or hours 
One man can't bargain for himself (collective 

Increases efficiency, better co-operation with 

management, men can work better 
Protects worker's job, gives him security 
Saves disputes, avoids disagreements, would 

be no more strikes 
Helps the workers 

If the unions are good ones (qualified answer) 
Don't know 

Part 2: Reasons Why All Workers Should Not Be Required to Join or 

Belong to a Union 

By membership 

Total homes Nonunion 

Should let worker decide for himself, leave it 
to the individual 

Places too much power in hands of union, 
unions are a racket 

It's not the American way, this is a free coun- 

Would cause a lot of trouble, lead to disagree- 

Just don't approve of unions 

Would depend on the union (qualified an- 

Takes away incentive to do better work, 
some men have better ability than others 


Don't know 















































* Less than .5%. 

Note: Percentages are based on the total sample of 1,000 people. 


"should be free to do as he wants, every worker should decide for himself, 
a right to do as he pleases'' Such a reply was given with the same fre- 
quency by individuals with as by those without union backgrounds. An- 
other 255 individuals objected to compulsory maintenance of member- 
ship on the grounds that "it's un-American, this is supposed to be a free 
country, this isn't a Dictatorship, against American principles 9 ' (23). 


The study described above was made in 1942 and reflects the attitudes 
of both union and nonunion workers and their wives. In Factory's 12th 
Survey of Workers Opinion, made in 1952, and involving a nation-wide 
survey limited to manufacturing wage-earners, the question was asked 
' 'Would you like to have a union shop (a rule that requires everyone to 
belong to the union) where you work? 9 ' Of workers in plants with union 
but without union shop, 43 per cent replied "Yes," and 42 per cent "No. 99 
When asked "Would you be willing to go on strike to force your employer 
to agree to a union shop? 99 41 per cent replied "Yes" and 38 per cent 
"No." According to Factory's summary, "there are still enough workers 
who are strongly pro-union shop to make a lot of strike trouble on this 
issue" (36, p. 86). 



As has been indicated in various parts of this volume most particularly 
under the discussion of bias of auspices in Chapter 12 the findings of 
management-sponsored attitude surveys are questioned by many union 
leaders. On the other hand, there seems to be considerable hesitancy on 
the part of union leaders to conduct surveys of their own unions, al- 
though, as indicated in a study by Davis and St. Germain (6a), such 
surveys can be useful in assessing policies and practices of the local 
union. One outcome of such reluctance, according to French, Korn- 
hauser, and Marrow, is a tendency for "the higher officials to become 
remote from the feelings of the rank and file ... Top union lead- 
ers/' it is added "are likely to emphasize strictly economic aims and 
to assume that the membership (even where this includes large numbers 
of workers entirely new to unionism) retains the same attitudes as those 
that prevailed during their own earlier days in the union" (12, p. 10). 

A "Pilot" Study of a Steelworkers' Union: The fact that there are few 
surveys which have been sponsored by unions adds interest to those which 
have been conducted in a union frame of reference providing information 
on why workers join unions. Among these is a "pilot study/' by Seidman, 
London, and Karsh (25), included in research conducted by the Indus- 


trial Relations Center of the University of Chicago. Subjects of the study 
were members and officers of a midwest local of the United Steelworkers 
of America. Interviews were conducted with three groups; viz., a leader- 
ship group, an active rank-and-file group, and an inactive rank-and-file 
group. In addition, observations were made of all types of formal and in- 
formal meetings and other union activities. In the leadership group, 28 
out of 36 men who satisfied the criterion for "belongingness" in this group 
(holding a union office, being a grievance man, a committee chairman, 
or being a past president of the local) were interviewed. Of the active 
members, 24 out of 43 were interviewed, an active member being defined 
as one who had attended between 4 and 7 union meetings in the past 
year. Finally a one-per-cent random sample was drawn from inactive 
union members, represented by those who had not attended any meetings 
in the year preceding the survey. Interviews were conducted with 62 out 
of 128 members constituting this sample. 8 In all, 114 interviews were 
conducted in the course of the survey. 

According to the investigators, the findings showed that the overwhelm- 
ing majority in all three groups had joined their union with some "degree 
of conviction." In the leadership group, the percentage was 86 per cent; 
in the active members group, 83 per cent; in the inactive group, the per- 
centage who had joined with some "degree of conviction" was 63 per cent. 
Most frequently mentioned reasons for joining were family background, 
earlier work- or union-experience, and personal experiences within the 
plant. Further analysis of the interviews revealed that the reasons given 
above were more influential in causing the active members to join than 
they were in causing the inactive members to join. Some members indi- 
cated that informal group pressures played a large role in their decision 
to join the union. Investigators state that available evidence suggests the 
hypothesis that membership in the union is in many instances a matter 
of social expediency, rather than the outcome of a logical analysis of the 

Approximately 95 per cent of the workers in the plants from which 
the subjects were drawn are union members. Those interviewed were 
asked what they thought were the reasons which kept the remaining 5 per 
cent of workers from joining the union. With respect to the question of 
whether these nonunion workers should be forced to join the union, 96 
per cent of the leadership group, 75 per cent of the active members, and 
54 per cent of the inactive group favored compulsory membership. Find- 
ings are interpreted as revealing that union members resent their non- 

8 Since the authors were primarily concerned with a qualitative analysis of the prob- 
lem being studied, they stopped interviewing when patterns of responses became 


union fellow employees and believe themselves justified in bringing pres- 
sure to bear on these nonconformists. 

A general conclusion drawn from this "pilot" study by the investigators 
is that reasons for joining unions are numerous and frequently without 
a logical basis. However, as for those who claim that workers join unions 
for economic reasons alone, the investigators report that not one of the 
114 members interviewed said that he joined the union in order to get 
higher wages! The latter finding, which is in complete contradiction with 
results of other studies conducted under a variety of auspices, naturally 
leads to speculation concerning the possibility that biasing influences, of 
the type discussed in Chapter 12, may have affected both the interview 
findings and the interpretation of results. 4 

A Comprehensive Survey of Members of an AFL Union: A compre- 
hensive survey of members' attitudes toward their union is found in a 
study conducted, in 1949, under the auspices of the Warehouse and 
Distribution Workers, Local 688, International Brotherhood of Team- 
sters, Chauffeurs, Warehousemen and Helpers of America, AFL, in St. 
Louis, Mo., with the union paying half the costs (24) . 

The study involved interviews patterned by a questionnaire which in- 
cluded 129 questions on a 13-page mimeographed schedule. Questions 
were "pretested" for reliability (understandability and unambiguity) and 
validity (relevance to the issue up for study) on small groups of union 
members, and two major revisions in the entire schedule were made as a 
result of the pretest. However, no statistical data as to reliability and 
validity are presented. 

A sample of the 4,100 members whose shops had been organized for 
at least seven years (of a total union membership of approximately 8,500) 
was accumulated by drawing every ninth name and address from an alpha- 
betical file of union members' names and addresses, with the omission of 
a small proportion of the members who lived in Illinois. 5 In general, a 
large number of the workers involved were in jobs classified as semiskilled. 
In all, 475 names were drawn and questionnaires were completed in the 
case of 392 of these in interviews conducted in the workers' homes. Ac- 
cording to Rose, there was insignificant bias in the sample as a result of 
the withdrawal of the names of men who had been fired, and probably 
more bias in the case of those who refused to be interviewed. Nevertheless, 
the sample is considered by the investigator to be generally representative 

4 One of the authors was at one time a member of the Amalgamated Lithographers 
of America and a past president of Local 6, UAW, CIO. Another had union experience 
with the UAW, CIO and with the new IUE, CIO. 

5 Although only shops which had been organized for seven years were canvassed for 
the sample, approximately 60 per cent of the sample could not date membership back 
for seven years. The study, according to the investigator, should "not be thought of as 
one of 'old union members/ " 


of union members in shops which had been organized for at least seven 

In addition to the chief investigator, interviewers included 50 under- 
graduate interviewers from Washington University who received a mini- 
mum of two hours training and followed a carefully outlined program in 
approaching union members. In articles carried in union newspapers 
before and during the study, the workers were told that the study was 
being conducted in their interests, had full union sanction, and that their 
responses would be treated anonymously by disinterested scientists and 
students. Each interviewer carried a copy of the union newspaper (Mid- 
west Labor World) which contained a write-up of the study plus an in- 
troductory letter signed by the union director. 

Of particular interest, for purposes of this section, are findings of the 
survey dealing with the question of why the workers joined the union. 
Replies to the question "Why did you join the union?' 9 are given in Table 
32. According to Rose, "apparently almost one-half of the members 
(those giving reasons 1 and 8) think of themselves as having joined the 
union involuntarily" (24, p. 60) . 

TABLE 32. Reason for Joining the Union (392 Members) (after Rose, 1952) * 

Reason given Percentage of members t 

1. "Had to I work in a union shop" 45.9 

2. "For my own benefit" (general, but personal) 20.9 

3. "It is a good cause" (general, but impersonal) 16.3 

4. "For higher wages" 7.7 

5. "For better working conditions" 6.6 

6. "For security" 4.3 

7. "There is strength in numbers'* 3.3 

8. "The majority wanted it" 2.8 

9. No answer 1.3 

* This and subsequent tables from the study by Rose are reproduced, with 
permission, from Union Solidarity, The University of Minnesota Press, 
(Copyright) 1952. 

t The figures add up to more than 100 per cent because a person could 
give more than one answer. 

A number of the items in the questionnaire dealt with the question 
of what the workers expected from the union, i.e. what they considered 
to be its major purposes. Included among such items was the question 
"What do you consider to be the purposes of your union? (What is a 
union for?) 99 This was treated as a free-answer question and interviewers 
did not prompt the respondents but allowed them to answer just as they 
pleased. In striking contrast with the findings reported in the case of the 
steelworkers union is the fact, shown in Table 33, that 75.3 per cent of 

Not italicized in the original. 


TABLE 33. Opinions of Union Members on the Purposes of the Union 

(after Rose, 1952) 

of members 

Purpose mentioned mentioning * 

"Get specific economic benefits" (higher wages) 75.3 

"Get job security" (including seniority) 31.1 

"Gain rights" (e.g., "welfare," free speech, fair deal) 16.6 

"Get benefits off the job" (recreational, medical, legal) 10.7 

"Organize labor; get solidarity for bargaining" 9.4 

"Raise the standard of living" 7.9 

"Make labor and management more co-operative" 5.1 

"Increase fellowship among workers" 3.8 

Miscellaneous; "Don't know"; no answer 2.6 

* The figures in this column add up to more than 100 per cent because a 
person could give more than one answer. 

a representative sampling of members of Local 688 consider the purpose 
of the union to be that of ''get specific economic benefits" in the form 
of "higher wages" An appreciably high percentage (31.1) also refers to 
the functions of the union in obtaining "job security" including seniority. 
By contrast, "make labor and management more co-operative" and 
"increase fellowship among workers" are viewed as purposes of this 
union by an extremely small percentage of the members. 

In the course of the interview, the question was asked "What do you 
think are the main things your union should work for right now, either 
through collective bargaining, or through social action in the commu- 
nity?" According to Rose, from an examination of replies to this question, 
as shown in Table 34, "there can be little doubt that most of the members 
see the main function of the union to be the traditional activity of col- 
lective bargaining to gain higher wages and better working conditions" 7 
(24, p. 151). 

As is apparent from Table 34, a certain proportion of respondents 
characterized by Rose as a "significant minority" -emphasized "recent 
supplementary functions of the union" in the way of securing security and 
health benefits, engaging in political action, etc. Attitudes towards these 
"extra" services were explored by another question: "Do you think that 
the union should put more time and money into getting higher wages 
from the employers or should it put more into a health plan, social and 
recreational activities, an insurance plan, and other things of that kind?" 
A comparison of Tables 34 and 35 shows inconsistencies in members' 
answers in that in reply to the second question the greater emphasis is 

T Not italicized in the original. 


upon "other things" rather than upon wages. According to Rose, the 
simplest explanation of this inconsistency appears to be that the spon- 
taneous replies in Table 34 are based partly on the traditional functions 
of the union which might naturally come to mind first in answering a 
question about the main purposes of a trade union. The weight assigned 
to supplementary functions when these are compared with traditional 
functions, such as getting higher wages, shorter hours, etc., "suggests that 
the members of this union feel a greater need" for the supplementary 
functions "but have not been fully educated to think in terms other than 
the traditional ones" (24, p. 143) . In a footnote to this explanation, Rose 
points out that "another factor to be considered is that many workers feel 
that their wages are as high as they can get now, but that they can get the 
supplementary benefits by union activity now" (24, p. 143). Considera- 
tion of the over-all situation at the time the study was made, and com- 
parison with findings from other studies, distinctly suggest the possibility 
that this is the prime rather than an alternative explanation of the facts 
presented by Rose. Involved here again are problems in the interpreta- 
tion of survey data which have been discussed at some length in Chap- 
ter 12. 

TABLE 34. Attitudes of Union Members Regarding the Union's Main 
Functions (392 Members) (after Rose, 1952) 

Question: "What do you think are the main things your union should -work 
for right now, either through collective bargaining or through social action 
in the community?" 


Answer of members * 

Higher wages and/or collective bargaining 33.7 

Better working conditions in the shop 18.4 
Benefits outside the shop, such as LHI t 9.4 

Job security and seniority 8.4 

Political action 5.4 

Organizing the unorganized 5.1 

Better employer-employee relationships 3.8 

General and community welfare 3.1 

Miscellaneous 6.1 

Nothing 4.9 

"Don't know" 19.1 
No answer 7.7 

* The percentages total more than 100 because some people gave more 
than one answer. 

t LHI is the Labor Health Institute, the free medical service organization 
which the union secured for most of its members through collective bargain- 
ing with the employers. 


TABLE 35. Relative Support among Union Members for "Extra" Union 
Services (392 Members) (after Rose, 1952) 

Question: "Do you think the union should put more time and money into 
getting higher wages from the employers or should it put more into a health 
plan, social and recreational activities, an insurance plan, and other things of 
that kind?" 


Answer of members 

"More into higher wages" 24.3 

"More into other things" 38.1 

"About same amount" 35.7 

"Don't know"; no answer 1.9 

Total 100.0 

Sources of "Union Solidarity": The detailed report of the extensive 
survey of Teamsters Local 688 presents many findings bearing upon the 
sources of "union solidarity." Following is a summary of a few facts and 
conclusions which are of particular interest in appraising the statements 
about the role of the union cited in the introductory section of this 

1. The fact that 93.4 per cent of the members replied "need a union" 
to the question "Do you think -you need a union to buck the employer 
for you, or could you do as well by yourself?" leads to the conclusion that 
all but a small percentage of members of Local 688 are convinced that 
a union is necessary, including those who joined the union "involun- 
tarily." According to Rose, "being forced to join the union because the 
company has a union shop is not a major deterrent to union solidarity" 
(24, p. 185). 

2. A number of questions asked in the survey, and also other data cited 
by Rose, have a bearing on the view, cited on page 334, that the union 
represents the prime medium for the "substitute" satisfaction of "psycho- 
logical" and "social needs" of self-expression (4); the "ties and bonds of 
group relationships" (13); "understanding the forces and factors at work 
in their world"; "integrity" (1), etc. Following are sample findings in 
this area 

(a) The "average" member attends about one-half of the total num- 
ber of monthly meetings held during the year. Only a small minority 
speak up at union meetings. 8 

(b) On the question of "democracy" in the union, 65.6 per cent be- 
lieve that important matters affecting the union should be decided by 
the "rank and file." Simultaneously, approximately 60 per cent are of 

8 The important problems of extent of member participation in union activities, fac- 
tors affecting participation, and also the influence of strongly organized minority groups 
are discussed by Goldstein (14a), Sayles (24a), and Strauss and Sayles (26c). 


the opinion that such matters are actually decided either by the union's 
director or the stewards' council. "Those whose desire for democracy is 
frustrated, because they believe that major union policies are set by the 
director or the stewards' council rather than by the rank and file, are 
relatively the poorest attenders at union meetings*' (24, p. 70). 

(c) Attitudes toward supplementary union activities in the way of 
educational and social activities provide a clue in the evaluation of the 
attitudes of members toward the union as a substitute for other social 
organizations. When asked "Would you like to take an education course 
now?", only 9.2 per cent answered "yes." When asked "In addition to 
having regular business taken up at the union meetings, 'would you like 
to have discussions, outside speakers, labor and political movies, or sing- 
ing?", approximately 43 per cent of respondents answered "nothing 
extra." Or the remaining 57 per cent, who wanted additional activities, 
the preference was for "discussions" and "outside speakers." The total 
percentage wanting parties, dances, entertainment, prayers, etc., was 3.1. 
According to Rose, "it cannot be said that union leaders have been 
successful in convincing the members that workers' education is an im- 
portant thing for them" (24, p. 102). A similar conviction appears to be 
lacking as to the role of this union as a focus of social relationships. 

(d) According to Krech and Crutchfield, one reason for the superior 
status of the union in furnishing "substitute" satisfactions is that "labor 
organizations provide the worker with opportunities to achieve positions 
of leadership and authority . . . opportunities that are denied him on 
the job" (19, p. 548). When members of Teamsters Local 688 were 
asked "What chance do you think you have for getting ahead and be- 
coming a foreman or shop owner yourself some day?" 48.7 per cent an- 
swered "No chance at all"; 17.4 per cent said "Fair chance" and 3.8 per 
cent, "Good chance" When asked "Have you ever thought of trying to 
become a union officer or staff member?" 83.3 per cent answered "No, 
never"; 10.4 per cent, "Yes, occasionally," and 5.3 per cent "Yes, often" 
A detailed analysis of such findings leads Rose to conclude that "The great 
majority of the members see the union as an instrument for achieving 
common goals. Only a few think of it as an avenue to leadership or per- 
sonal promotion" (24, p. 150). The analysis also shows that those who 
think of trying to become union officials are also the ones who are more 
likely to think that they have opportunities for promotion in the shop. 
Motivation here appears to have an internal source in the individual's 
belief in his own capacity rather than in the nature of the social or- 

(e) No differences were found between churchgoers and non-church- 
goers with respect to attendance at union meetings and loyalty to the 
union. According to Rose, this means "that there is no conflict between 
these two institutions in the minds of their common members. . . . 
Rather, they seem to consider each demand on their time and loyalty 
in terms of its intrinsic merits" (24, p. 192). 

3. The survey of Teamsters Local 688 also considered the question of 
"whether a high degree of union solidarity is associated with antagonism 
toward the employer" whether, in other words, there is "a 'class struggle' 


in the minds of the workers so that the large number of pro-union workers 
are anti-employer" (24, p. 65). 

(a) In reply to the question "Do you think your foreman is fair to the 
workers?" 72.2 per cent of all members answered "Very fair most of the 
time," and 21.9 per cent "Sometimes fair, sometimes not fair" 

(b) In reply to the question "Do you think the top officials and bosses 
of the company are fair to the workers?" 50 per cent of all members re- 
plied "Very fair most of the time"; 27.8 per cent, "Sometimes fair, some- 
times not fair"; only 4.8 per cent replied "Hardly ever fair" 

(c) Members did not anticipate that the union will always be able to 
get wage increases. Furthermore, no relationship was found between 
expectation as to wage increases and whether or not the workers think 
that the employer is fair. 

(d) Consideration of the findings summarized above leads Rose to 
conclude that "The evidence which arises from this study seems to sug- 
gest that there is no inflexible antagonism toward employers on the part 
of the large majority of workers. If anything, the evidence supports the 
opposite contention that the union movement is a buttress for the free 
enterprise system" (24, p. 66). 

Consideration of the theoretical implication of findings bearing upon 
attitudes of union members toward employers leads Rose to the further 
generalization that: "People can have loyalty to two (or more) groups 
or two sets of values, even wherLthese groups or values are in conflict. In 
concrete terms, loyalty to the union does not mean disloyalty to the em- 
ployer? In this study/' it is added, "we have not examined the concrete 
psychological mechanisms by means of which this dual loyalty can occur, 
but this would be a worth-while subject for further research" (24, p. 

Motives for Joining Unions an Evaluation 

A comparison of the studies described in preceding pages shows that 
findings and conclusions bearing upon reasons for joining unions, and 
upon members' perception of the unions' goals, are far from uniform. 
Here again, as in the evaluation of attitudes toward pay, security, job 
status, etc., variations in survey methods, in periods during which the 
surveys were conducted, biases of various kinds (including those of inter- 
pretation), etc., contribute to the differences cited above. Nevertheless, it 
appears possible to point to the following as trends characterizing the find- 
ings of representative surveys bearing upon the attitudes of employees 
toward unions: 

1. It is clearly apparent that economic needs expressed in the desire 
for higher wages, job security, and fringe benefits play a highly signifi- 
cant role in leading workers to join unions. A major conclusion from the 

Not italicized in the original. 


study of Teamsters Local 688, that "the strong loyalty to the union is 
closely associated with ... the success that the union has in achieving 
its goal of increasing the workers' income (and) security" 10 (24, p. 183) 
appears to have very wide, if not universal application. There seems 
no question, as suggested by Tripp, that "the psychology of insecurity, 
awareness of job scarcity, provides a basic drive in the formation of 
unions and in their subsequent activities to protect and advance job 
interests" (27, p. 95). 

2. The fact that a large proportion of members describe their mem- 
bership as "involuntary" or refer to "compulsion" indicates that fear has 
been a factor in the increase of union membership. It is evident that there 
are social forces at work in the growth of union membership, but survey 
data suggest that these are as much the social forces of group pressures 
and intimidation, as social motives seeking gratification in the "ties and 
bonds of group relations and community life" (13, p. 7). 

3. As is apparent from the surveys of younger people, social forces are 
also operating through the family group, as indicated by the fact that 
approval of the union and pro-labor attitudes are more common among 
children of parents who belong to unions than among other children. Of 
great significance are social norms associated with socio-economic status. It 
cannot be assumed, however, that inherent in the operation of such norms 
is the perception of the union as a prime medium for self-expression, for 
the gratification of the needs of self-respect, creativity, etc. In fact, the 
frequent reference, in a number of surveys, to distrust of union leaders 
and to nondemocratic practices in union administration may well be 
interpreted as indicating that such needs act as a deterrent to voluntary 
union membership. On the other hand, survey findings indicate that fear 
listed above as a cause for joining unions may also operate against 
voluntary membership by employees who, in spite of the protective cloak 
of labor legislation, are still afraid of reprisals on the part of their em- 

Reference to trends noted above is not included to suggest that they 
can invariably be generalized to any particular situation. The clamor for 
gratification of the needs for self-expression and a variety of social needs 
may loom large in the organization of a union in a small town dominated 
by a powerful single industrial corporation which has provided little or 
nothing in the way of social and recreational facilities. Similar needs may 
prove to play a minor role in leading workers to join unions in larger cities 

10 This quotation is abstracted from a summary of "practical conclusions" by Rose 
who writes: "The strong loyalty to the union is closely associated with two variables: 
(a) the success that the union has in achieving its goals of increasing the workers' in- 
come, security, and job satisfaction; (b) the amount of participation in union activi- 
ties the rank and file members engage in" (24, p. 183). 


with diversified industries and a variety of cultural and recreational facili- 
ties even though, having joined, the worker will respond to the influence 
of the group and develop bonds of companionship which can have an 
important bearing upon future attitudes and behavior. 

It may generally be true, as Rose suggests in the discussion of Teamsters 
Local 688, that members develop a sense of solidarity as a result of ex- 
perience with the union (24, p. 61 ) . However, again, the conclusion with 
respect to Teamsters Local 688, that "members went in without knowing 
much about the union and having relatively neutral attitudes toward it" 
(24, p. 61) appears to have wide application to workers in general. 

To state this generalization is not to deny that in many instances 
"worker solidarity . . . has come to mean more than advantages to be 
exacted from the employer" (24, p. 4); that for some workers it has come 
to mean co-ordinated social action in a broad range of activities, including 
political action. The delineation of trends covering motives in joining 
unions is not intended to deny that, in some instances, unions are viewed 
by individual workers as a means of gaining recognition and status and 
a behavior role which cannot be achieved in the plant. There is no intima- 
tion that unions need to be or are necessarily always so organized as to 
make impossible the very kind of participation which is needed for the 
gratification of those "psychological" and "social" needs which, it is 
claimed, it is the union's goal to satisfy. The description of activities in the 
business quarters of the newly formed UAW, CIO local of the Ford plant, 
in 1941, provides concrete evidence that the office of a particular union 
can be "the friendliest place" a worker may know to visit for obtaining 
advice on all kinds of problems (29) and that unions can and perhaps 
do in some instances "perform the same kind of function that the political 
ward bosses did a generation ago" (24, p. 11). It may also well be, as 
Knickerbocker has suggested, that as collective bargaining becomes an 
increasingly accepted process, this function of the union will be de- 
emphasized, and what are now regarded by members as supplementary 
or secondary functions will be recognized as primary functions of the 
union (18). 

Recognition of this possibility immediate or remote as it may be 
does not, however, necessitate the view that unions do all of these things 
referred to above better than does the individual industrial plant or than 
industry collectively. In particular, exception can be taken to the view 
taken by Krech and Crutchfield that the union can better satisfy the 
needs for participation, for self-expression, for self-respect, for status, 
etc., than do other types of social organizations. Moreover, whatever may 
be the outlook for the future, many unions fail to provide opportunities 
of the type referred to above fail to "offer to their members an exultant 


sense of progress toward an increasing share in controlling their condi- 
tions of work" (19, p. 549). As Krech and Crutchfield suggest, some 
unions can be and are dictatorial in their procedures and "are not adverse 
to using violence and 'goon squads' to silence their members who would 
participate in free speech" ( 19, p. 549); others operate as a tight bureauc- 
racy, with opportunity for leadership restricted to the self-selected few; 
the policies and practices of still others reflect the drive for continued 
power by their leaders, rather than an urge to render services and op- 
portunities of the types described above. 

Such considerations make it difficult to accept the point of view that 
"most unions offer the individual worker roundabout, constructive, 
healthy solutions to the blockage of many needs blockages that he can 
trace either directly or indirectly to his job" ( 19, p. 5 f.) to a greater extent 
than do religious or other types of organization, including the progressive 
company with an extended program of employee activities. Under any 
circumstances, there is the fact, evident from the discussion of attitude 
surveys on earlier pages of this text, that when provided with an oppor- 
tunity to tell what is important to them, workers refer to the union only 
infrequently. Young people and employees alike fail to show an over- 
whelming desire to join unions, or to be forced to stay in them, although 
they want the freedom to join. 

On the other hand as the record shows workers in the United States 
have joined unions in increasing numbers. Certainly government support 
of the union movement directly through labor legislation and indi- 
rectly through other means has contributed to this increase in union 
membership (34, 35). As indicated earlier, management abuses of its 
power have contributed to this development as has also energetic leader- 
ship within the union movement. Other forces including perhaps those 
implied in Lewin's concept of the "marginal man" whose "span of free 
movement" is limited as a nonunion employee in a union shop (21) 
cannot be overlooked in considering the growth of unionism in the United 
States. Whatever the reasons for joining unions, the fact remains that 
when faced with the direct issue in the form of a vote, large proportions 
of the workers in many plants have voted for the union. Thus, in a survey 
of elections held under the auspices of the National Labor Relations 
Board in the area of Buffalo, New York, Sheldon found that "in only 
two of the 127 elections studied did a majority of the eligible votes 
fail to authorize a union shop. Over 96 per cent of the workers who voted 
in the election cast 'yes' ballots. More than 91 per cent of those who were 
eligible to participate did vote" ( 26, p. 37) . In the case of Teamsters Local 
688, Rose reports that "the proportions in favor of the union in NLRB 
elections ranged from 94 to 98 per cent for the various shops" (24, p. 60) . 


Not all surveys show such predominating support of the union in 
NLRB elections. Thus, according to a survey reported by Fisher, Rudge 
and Neblett, Inc., in 1952, "employers won a third of the elections held 
during the second quarter of 1952 the employees choosing a union 
in 1,365, or 67 per cent of the 2,020 contests in that period, as compared 
with an average of 69,9 per cent in the previous quarter and 74 per cent 
for the fiscal year 1951. Moreover, in the quarter referred to, 23 of the 34 
decertification elections held resulted in upsetting of the union" (37, p. 
4) . An interesting sidelight in the evaluation of these data is found in the 
fact that the percentage of workers voting in such elections has grown 
from 52 per cent in 1948 to 86.6 per cent in the April-June period of 

In spite of variations in trends noted above, and of reports showing a 
"levelling-ofF ' of union members